2026-08-16 - 2026-11-16
Overview
3 Releases published by 1 user
Published
v1.0.0-rc3
Published
v1.0.0-rc2
Published
v1.0.0-rc1
164 Pull requests merged by 1 user
Merged
#457 Add snapshot files to list a snapshot's files without restoring it (closes #435)
Merged
#456 Refuse an exclude pattern that does not compile at config load
Merged
#451 Print a failed command's error once on stderr (closes #433)
Merged
#454 Name the S3 keys when storage_url moves to s3:// first (closes #428)
Merged
#452 Name the setting and value when a size does not parse (closes #429)
Merged
#453 Refuse a relative index_path at config load (closes #430)
Merged
#442 Skip vaultik's own index and PID file in a backup (closes #421)
Merged
#444 Report the Go build info version when linker flags set none (closes #422)
Merged
#450 Do not log an interrupted blob upload as an error (closes #434)
Merged
#448 Say plainly when the private key cannot decrypt a snapshot (closes #427)
Merged
#443 Refuse config keys that no setting has (closes #419)
Merged
#449 Report an unreadable blob once, without its bytes (closes #424)
Merged
#445 Bind the local index only after a backup writes the marker (closes #420)
Merged
#446 restore and verify report an unknown snapshot ID as not found (closes #426)
Merged
#447 Exit 1 when a deletion prompt is not confirmed (closes #423)
Merged
#417 Shorten internal/vaultik tests under the race detector (closes #416)
Merged
#441 Keep only open work in docs/TODO.md (closes #440)
Merged
#415 Format Markdown with prettier and move documents out of the root (closes #411)
Merged
#412 Label the version report date line "commit date"
Merged
#414 Print remote info and verify status lines through internal/ui
Merged
#413 Write the apps snapshot in config init only on macOS
Merged
#407 Correct two false doc comments (closes #403)
Merged
#396 Do not prune when a backup misses an earlier mount point (closes #395)
Merged
#393 Delete metadata directories a backup left without a manifest (closes #390)
Merged
#394 Do not prune after a backup that found a configured path empty (closes #388)
Merged
#392 Go on with a backup when the destination marker cannot be checked (closes #389)
Merged
#391 Keep unreferenced blobs newer than prune_grace_period
Merged
#386 Record the defect-hunt fixes in TODO.md Completed Steps
Merged
#385 Do not prune after a backup that skipped unreadable files
Merged
#381 Exit non-zero when snapshot create --skip-errors skips anything
Merged
#384 Keep the local index when remote nuke cannot list or find the destination
Merged
#383 Stop counting directories and unchanged symlinks as files to back up
Merged
#349 Refuse a backup destination an unplugged volume left empty
Merged
#378 Warn about restored ownership only when an entry's owner or group differs
Merged
#375 Make snapshot list report a changed destination instead of advising prune
Merged
#342 Count a file as deleted only if the previous snapshot of its name had it
Merged
#377 Name each prefix once in restore, verify and purge errors
Merged
#352 Refuse credentials in a storage URL's endpoint or rclone remote name
Merged
#380 Run $EDITOR through the shell in config edit
Merged
#370 Fail snapshot remove when a destination-only snapshot's metadata cannot be deleted
Merged
#379 Label the index cleanup totals as totals
Merged
#372 Free a cached blob once no file still to restore needs it
Merged
#376 Keep the whole startup error when it contains "): "
Merged
#369 Silence prune and snapshot removal status under --cron and --quiet
Merged
#374 Read the rclone config file named by RCLONE_CONFIG
Merged
#367 Make **/ in an exclude pattern match zero directories
Merged
#373 Make the size defaults and the chunk_size minimum 1000-based
Merged
#371 Keep remote metadata when purge cannot remove a snapshot locally
Merged
#365 Fail restore for a requested path that matches nothing
Merged
#368 Make remote nuke delete every object under metadata/ and blobs/
Merged
#357 Make config set create a list for an index under a missing key
Merged
#331 List a file:// destination through a symlinked directory
Merged
#362 Report blob counts from snapshot verify --deep --json
Merged
#354 Limit and retry the module download in the Docker build
Merged
#366 Record each snapshot's uncompressed size when its backup runs
Merged
#353 Stop restore --verify re-reporting files that failed to restore
Merged
#341 Record the stored size of a blob a backup finds already stored
Merged
#364 Return errors instead of exiting on rollback, rows close and blob temp file cleanup
Merged
#363 Stop a stalled snapshot database read when restore or deep verify is interrupted
Merged
#336 Make prune and remote nuke fail when a delete fails
Merged
#333 Make an S3 listing the server cuts short an error
Merged
#348 Report blobs shallow verify could not check apart from missing ones
Merged
#361 Restore the other files when a chunk is in no blob under --skip-errors
Merged
#359 Record each snapshot's new chunk size when its backup runs
Merged
#343 Exit as interrupted when Ctrl-C lands during a local index write
Merged
#358 Print snapshot list and purge sizes like every other command
Merged
#355 Stop a stalled blob read when restore or deep verify is interrupted
Merged
#350 Make a mistyped subcommand under a group command an error
Merged
#335 Make a leading **/ in an exclude pattern match at the top level
Merged
#330 Refuse a relative file:// storage_url and decode its escapes
Merged
#347 Remove the blob temp file when a backup stops part-way
Merged
#346 Check --keep-newer-than before the backup starts
Merged
#344 Stop remote info counting a non-blob file under blobs/
Merged
#339 Make snapshot remove fail for an unknown snapshot ID
Merged
#329 Bind the local index to the s3.* destination too
Merged
#328 Reject a retention duration longer than about 292 years
Merged
#287 Stop a cancelled backup at once under --skip-errors
Merged
#283 Make remote nuke delete leftover .partial uploads
Merged
#285 Never skip a local index error for a directory or symlink
Merged
#282 Count a file a backup could not store as failed
Merged
#279 Report unknown blob figures for a snapshot whose manifest cannot be read
Merged
#278 Keep the progress line of a multi-path snapshot within 100%
Merged
#276 Document that --json skips the confirmation prompt of snapshot remove and prune
Merged
#277 Give a second snapshot create in the same second its own ID
Merged
#275 Stop a backup on a symlink whose target cannot be read
Merged
#274 Write rclone uploads under a temporary name and move them into place
Merged
#273 Exit 130 and say so when a command is interrupted
Merged
#264 Correct doc and help sentences that are false about the code
Merged
#265 Speed up the vaultik and database package tests
Merged
#263 Fix two misleading messages: the readable-config warning and the purge listing error
Merged
#262 Pass s3.part_size to the multipart uploader
Merged
#261 Read the snapshot name using the stored hostname
Merged
#259 Leave remote info orphan figures unknown when a manifest is unreadable
Merged
#260 Quote a string setting that YAML would read as a number
Merged
#257 Store and compare file mtimes to the nanosecond
Merged
#258 Make the process-wide lock atomic with flock
Merged
#256 Write only the document to stdout from snapshot remove --json
Merged
#254 Count each file, byte and upload once in backup statistics
Merged
#252 Keep command output to the README's stdout and stderr rules
Merged
#255 Start the progress reporter once per snapshot, not once per path
Merged
#250 List the files under a path by path, not by string prefix
Merged
#249 Return an error, not a panic, on a malformed snapshot database
Merged
#248 Join the S3 prefix to every key with one slash
Merged
#245 Apply restored owners, modes and times in an order that keeps them
Merged
#247 Mark github.com/spf13/pflag as a direct dependency in go.mod
Merged
#244 Load a config that has no age recipient
Merged
#243 Skip files of an unreadable blob under restore --skip-errors
Merged
#236 Re-vendor the canonical files from sneak/prompts at dd4027b
Merged
#242 Run the local index in WAL mode with a busy timeout
Merged
#241 Record the real uid and gid of backed-up files
Merged
#238 List a missing file:// destination directory as an error
Merged
#237 Reject a retention duration with characters outside its number-and-unit parts
Merged
#234 Re-chunk a known file whose chunks no uploaded blob holds
Merged
#212 Stamp the tag or short commit in a plain docker build (closes #211)
Merged
#209 List only after-1.0 work in the README roadmap
Merged
#210 Run the disk-full restore test again
Merged
#205 Document that migrations are supported and none are added before 1.0 (closes #68)
Merged
#114 next: integrate accumulated work into main
Merged
#202 Vendor the canonical .golangci.yml (gomodguard_v2)
Merged
#201 Route direct-stdout command output through internal/ui (closes #149)
Merged
#200 Mark a snapshot complete only after its metadata export succeeds
Merged
#199 Correct the security claims in docs and comments, and record the accepted risks
Merged
#198 Reject a metadata database truncated to the age header and nonce
Merged
#197 Bound download expansion and escape control chars on the terminal (closes #164)
Merged
#195 Validate blob hashes, offsets and lengths read from the destination
Merged
#196 Parse the age identity key once and accept every identity in it
Merged
#194 Reject a decrypted snapshot database that is not the requested one
Merged
#193 Check blob sizes and the database in shallow verify
Merged
#192 Add negative and boundary tests for blobgen and types (closes #170)
Merged
#189 Wait for the interrupted operation to clean up before exit
Merged
#191 Remove the unused crypto path and write the blob-ID hash step once
Merged
#190 Scrub example config of real credentials and internal hosts
Merged
#188 Give the local index and its export copy an explicit 0600 mode
Merged
#187 Parse age_recipients at config load and never echo the entry
Merged
#186 Open the downloaded snapshot database read-only, on a private temp dir
Merged
#185 Abort the run when packing fails, even under --skip-errors
Merged
#184 Stop config set echoing secrets; reject credential-bearing storage URLs
Merged
#183 Reconcile purge against remote by hashed key, not human ID
Merged
#182 Restore files at 0600 and make the blob hash check unskippable (closes #163)
Merged
#181 Reject a blob_size_limit below the largest possible chunk (closes #167)
Merged
#180 Fail closed on unreadable manifests instead of losing blobs
Merged
#178 Default a scheme-less s3.* endpoint to TLS
Merged
#179 Scope the PID lock to mutating commands
Merged
#176 Keep restore writes inside the target directory
Merged
#175 Trust only uploaded blobs for deduplication (closes #148)
Merged
#173 Add fault-injection tests for interruption and corruption (closes #72)
Merged
#145 Quiet only the stdout UI under --json, not the log level (closes #112)
Merged
#174 Reconcile docs with code across the accuracy-sweep items (closes #74)
Merged
#147 Accept a remote key for restore and verify, and document it (closes #124)
Merged
#144 Add tests for internal/storage URL parsing and the file backend (closes #66)
Merged
#141 Stamp Docker image version from the host; flush profiles on error exit (closes #75)
Merged
#146 Reconcile the schema/migration docs with the code (closes #68)
Merged
#139 Report a prune count that could not be read as unknown, not 0 (closes #96)
Merged
#143 Write file:// blobs atomically via temp file and rename (closes #130)
Merged
#142 Map s3 not-found to storage.ErrNotFound in Get and Stat (closes #129)
Merged
#140 Use one duration parser and fix the --older-than months example (closes #123)
Merged
#135 Drop the lint-guard shell scanner, keep the Dockerfile.lint checks (closes #121)
Merged
#138 Hash the plaintext, not the encrypted bytes, in verify --deep (closes #131)
Merged
#132 Hash-verify the Go toolchain in the release workflow (closes #105)
Merged
#136 VACUUM snapshot metadata through the sqlite driver, not a CLI (closes #120)
Merged
#134 Delete dead code and stale fixtures, fix config set reindent
Merged
#128 Correct remote layout and privacy docs for hashed snapshot keys (closes #67)
Merged
#127 Trigger CI on next, not only main
Merged
#126 Correct --cron flag help to name warnings as unsuppressed (closes #87)
5 Pull requests proposed by 1 user
Proposed
#203 next: integrate accumulated work into main
Proposed
#399 Rewrite README.md and the other prose documents sentence by sentence (closes #397, closes #398, closes #431)
Proposed
#404 snapshot remove accepts a remote-only snapshot's remote key (closes #400)
Proposed
#405 Honor NO_COLOR in the terminal log handler (closes #402)
Proposed
#406 Missing-key errors for restore and deep verify name both key sources (closes #401)
180 Issues closed from 1 user
Closed
#435 proposal: list the files in a snapshot without restoring it
Closed
#455 An exclude pattern that does not compile is skipped during the backup instead of refused when the config loads
Closed
#433 cosmetic: a failed command prints its error twice on stderr
Closed
#428 cosmetic: config set storage_url to s3:// is refused until the S3 keys are set
Closed
#429 cosmetic: a non-numeric chunk_size or blob_size_limit fails with an error that names neither the setting nor the value
Closed
#430 cosmetic: index_path accepts a relative path
Closed
#421 The default config backs up vaultik's own live local index and lock file
Closed
#422 A binary from go install reports version dev although it was built from the tagged release
Closed
#434 cosmetic: Ctrl-C during a backup logs an upload error before the interrupted message
Closed
#427 cosmetic: the wrong-private-key error says "no identity matched any of the recipients" with a doubled prefix
Closed
#419 A misspelled or unknown config key is accepted silently
Closed
#424 cosmetic: a corrupted blob's raw bytes are printed in every error, once per file under --skip-errors
Closed
#420 A backup that fails before writing anything binds the local index to the configured destination
Closed
#426 cosmetic: restore and verify of an unknown snapshot ID report "object not found" on a hashed path
Closed
#423 cosmetic: a cancelled confirmation prompt exits 0
Closed
#416 internal/vaultik tests take 36-56 seconds and pass the 90-second timeout when the host is busy
Closed
#440 TODO: remove work logs; keep only open work
Closed
#418 Acceptance test: use vaultik end to end as a user before a release candidate
Closed
#411 Format Markdown with prettier and move documents out of the root, as REPO_POLICIES.md requires
Closed
#409 vaultik version labels the commit date "build date"
Closed
#410 snapshot verify and remote info print status lines without internal/ui
Closed
#408 config init writes a config whose first backup on Linux stops on /Applications
Closed
#403 Two code comments describe behaviour that no longer exists
Closed
#395 A backup while a volume mounted below a configured path is not mounted deletes the older snapshots holding its files under --prune
Closed
#390 prune never removes the encrypted database a backup left on the destination when its manifest upload failed
Closed
#388 A backup of a source directory left empty by an unmounted volume deletes every older snapshot under --prune
Closed
#389 The documented PUT-only source host fails every backup after its first: the destination marker check needs HEAD permission on S3
Closed
#387 A prune on one host deletes blobs of another host's backup in progress; that backup reports success and cannot be restored
Closed
#334 Record the defect-hunt fixes in TODO.md Completed Steps
Closed
#382 snapshot create --skip-errors --prune deletes the previous snapshot, the only one holding the skipped files
Closed
#310 snapshot create --skip-errors exits 0 and prints "Finished successfully." after skipping files
Closed
#292 remote nuke against a destination it cannot list deletes the local index first
Closed
#305 Directories and symlinks are counted as files to back up on every run, so an unchanged tree reports data backed up
Closed
#239 How should vaultik tell an unplugged file:// volume from a new, empty destination?
Closed
#316 Every restore by a non-root user warns that ownership was not preserved, even when it was
Closed
#317 After storage_url changes, snapshot list advises prune, which then refuses
Closed
#304 The "deleted" figure of a backup counts files that were not deleted since the previous snapshot of that name
Closed
#327 Restore and verify error lines repeat a prefix, and deep verify names S3 for any destination
Closed
#322 Credentials can still be put in storage_url, inside the endpoint or an rclone remote's parameters, and are then printed
Closed
#315 config edit fails when $EDITOR carries arguments
Closed
#340 snapshot remove exits 0 when deleting a destination-only snapshot's metadata fails, and its README promise of a prune command is false on several paths
Closed
#306 The index cleanup before each backup reports total rows as "remain"
Closed
#324 A restore of selected paths keeps decrypted blobs in the temporary cache until it ends
Closed
#320 A config error that contains "): " loses its start, including which setting is wrong
Closed
#294 --cron and --quiet do not silence the status lines of prune
Closed
#321 RCLONE_CONFIG is ignored, so an rclone destination configured there is not found
Closed
#338 Exclude patterns /**/x and a/**/b do not match as in gitignore
Closed
#313 Byte-size settings: the defaults and the chunk-size minimum are 1024-based while the docs, the parser and the messages are 1000-based
Closed
#337 snapshot purge exits 0 and still deletes the remote metadata when it cannot remove a snapshot from the local index
Closed
#309 snapshot restore ignores a requested path that matches nothing, and exits 0
Closed
#291 remote nuke leaves a snapshot database without a manifest, and non-blob files under blobs/, and says the store is empty
Closed
#308 config set with a list index on a key that does not exist writes a map, and then no command can load the config
Closed
#296 A symlinked metadata/ or blobs/ directory in a file:// destination lists as empty, and prune then deletes every blob
Closed
#314 snapshot verify --deep --json reports "missing": 0 and "verified": 0 when it fails
Closed
#351 The Docker build hangs for as long as a stalled connection to the Go module proxy lasts
Closed
#360 snapshot list's UNCOMPRESSED SIZE of an older snapshot changes after a later backup
Closed
#323 restore --skip-errors --verify reports a file that failed to restore again as a checksum failure
Closed
#300 A reused blob is recorded with the size of a local copy that was never stored, so snapshot verify fails on a good snapshot
Closed
#345 internal/database and the blob upload step still end the process on a rollback after Ctrl-C, a failed close of query results, or a failed temp file cleanup
Closed
#356 Interrupting a stalled read of the snapshot database leaves decrypted data in the temp directory
Closed
#290 prune and remote nuke exit 0 when a delete fails, and nuke still says the store is empty
Closed
#297 An S3 listing the server marks truncated without a continuation token is taken as complete, so prune can delete blobs still in use
Closed
#318 Shallow snapshot verify reports any error reading a blob's size as "Missing", and drops the error
Closed
#326 Under --skip-errors, one file whose chunk is in no blob aborts the whole restore before anything is restored
Closed
#311 snapshot list shows a NEW CHUNK SIZE of 0 B for a snapshot whose change is a modified file
Closed
#301 Ctrl-C during a backup can end with "FATAL: failed to rollback transaction" and exit 1 instead of the interrupted exit
Closed
#312 snapshot list and the snapshot purge listing print 1024-based sizes labelled MB, unlike every other command
Closed
#325 Interrupting a restore during a slow blob read leaves the decrypted snapshot database and decrypted file data in the temp directory
Closed
#319 A mistyped subcommand under snapshot, config, remote, database or completion exits 0 and does nothing
Closed
#303 Exclude patterns starting with **/ miss a match directly under the source path
Closed
#299 A relative file:// path is accepted, so the destination moves with the working directory and a backup can exit 0 and be unrestorable
Closed
#302 A failed or interrupted backup leaves its half-written blob in the temp directory
Closed
#307 An invalid --keep-newer-than is checked only after the backup, which is then reported as failed and retention never runs
Closed
#295 remote info counts a file under blobs/ that is not a blob as an orphaned blob
Closed
#293 snapshot remove reports success for a snapshot that does not exist
Closed
#332 A file:// destination whose metadata/ or blobs/ is a symlink to a missing target lists as empty, so prune deletes every blob
Closed
#298 With the s3.* configuration, a changed bucket or prefix is not detected, and the next backup reports success but cannot be restored
Closed
#289 A retention duration of 293 years or more deletes every snapshot
Closed
#288 Defect hunt on next before the release candidate
Closed
#204 Daemon mode after 1.0: watch for file changes so frequent backups skip the full scan
Closed
#286 An interrupted --skip-errors backup keeps opening every remaining file and logs each one as failed
Closed
#281 remote nuke reports the destination store empty while a leftover .partial upload remains
Closed
#284 Under --skip-errors, a local index error while recording a directory or symlink is skipped instead of stopping the backup
Closed
#280 A file skipped while a backup is processed is reported as "unchanged" and its bytes as "backed up"
Closed
#272 remote info shows 0 blobs and 0 B for a snapshot whose manifest cannot be read
Closed
#271 Progress lines over-report on the second and later paths of a multi-path snapshot
Closed
#268 The help and README do not say that --json skips the confirmation prompt of snapshot remove and prune
Closed
#270 Two snapshot create runs of one name within the same second fail on the snapshot ID
Closed
#269 A symlink whose target cannot be read is left out of the snapshot silently, even without --skip-errors
Closed
#266 A blob upload to an rclone destination cut off by a hard kill leaves a partial object that the next backup trusts
Closed
#267 An interrupted snapshot create, restore or verify exits 0 and prints nothing
Closed
#233 Doc and help sentences that are false about the current tree
Closed
#235 The internal/vaultik and internal/database tests take longer than the 20-second target
Closed
#240 Two misleading messages: a credentials warning for a config with no credentials, and a doubled error prefix
Closed
#232 s3.part_size is documented and accepted but has no effect
Closed
#230 A hostname containing _ makes per-name retention silently do nothing
Closed
#228 remote info counts a live snapshot's blobs as orphaned when a manifest cannot be read, and prints store names unescaped
Closed
#229 config set cannot store a string that looks like a number
Closed
#226 A file rewritten within the same second as its recorded mtime is never backed up again
Closed
#227 The process lock is check-then-write, so two writers started together can both acquire it
Closed
#251 snapshot remove --json prints its remote-cleanup warning on stdout, ahead of the document
Closed
#225 Backup statistics are wrong in the printed summary and in the snapshots table
Closed
#224 Command output breaks the README's stdout and stderr rules in three places
Closed
#253 A backup without --cron panics when a snapshot has more than one path
Closed
#223 Path filters match by case-insensitive string prefix: restoring /home/u/doc also restores doc2, Doc3 and doc.txt.bak
Closed
#231 A malformed snapshot database crashes restore with a Go panic instead of an error
Closed
#222 With an s3://bucket/prefix URL, objects are stored as prefixblobs/... instead of prefix/blobs/...
Closed
#219 Restore applies directory modes and times before writing the contents, and file modes before ownership
Closed
#246 The pre-commit hook fails on every commit: go.mod marks a directly imported module as indirect
Closed
#221 The documented "restoring on another machine" steps fail at the first command
Closed
#218 snapshot restore --skip-errors aborts on a missing or damaged blob instead of skipping the files that need it
Closed
#213 Re-vendor the canonical files from sneak/prompts at dd4027b
Closed
#217 The local index ignores its WAL and busy-timeout settings, so a read-only command can abort a running backup
Closed
#216 File ownership is never recorded, so a restore as root makes every file root:root
Closed
#220 A file:// destination that is missing (an unmounted volume) is treated as present and empty
Closed
#215 A retention duration with unparsed characters is misread: --keep-newer-than 1.0y deletes every snapshot
Closed
#214 A backup reports success but cannot restore a changed file after the snapshot holding its current version is dropped
Closed
#211 docker build . stamps dev, not the git tag or short commit
Closed
#125 Work remaining before vaultik 1.0
Closed
#208 README roadmap and TODO.md still list finished 1.0 work as remaining
Closed
#207 The disk-full restore test is still skipped although the fix it waited for has landed
Closed
#206 Run every build, test and lint step in the Dockerfiles at nice -n 19
Closed
#68 Reconcile the schema/migration story across code and four documents
Closed
#94 Decide whether feature/daemon-mode is 1.0 scope
Closed
#133 CI runner mac-studio fails every job before the first step (docker network pools exhausted, disk full)
Closed
#113 Run all linting in Docker via Dockerfile.lint + script/lint
Closed
#155 Check blob hashes, offsets and lengths read from the destination before using them
Closed
#156 Restore and deep verify must check the decrypted database is the snapshot that was asked for
Closed
#159 Ctrl-C during restore exits before cleanup, leaving decrypted data in the temp directory
Closed
#177 Interrupted metadata export leaves a snapshot silently treated as complete but unrestorable
Closed
#90 golangci-lint deprecation: gomodguard replaced by gomodguard_v2
Closed
#119 Review toward 1.0: identify and file all next steps
Closed
#73 Security review of the crypto and blob-generation seams
Closed
#76 Decide on renaming internal/log, internal/crypto and internal/types (revive var-naming)
Closed
#101 REPO_POLICIES.md mandates a 30s test timeout that is too tight to be safe
Closed
#149 Route direct-stdout command output through internal/ui
Closed
#171 Correct the security claims in the docs and comments, and write down the accepted risks
Closed
#152 A stream cut right after the age header reads as valid and empty, and restore then reports success
Closed
#164 Bound what a downloaded object may expand to, and escape control characters from the destination before printing
Closed
#165 Parse the age private key once, accept every identity in it, and stop showing it typed on the command line
Closed
#169 Shallow snapshot verify ignores object sizes and reports more than it checked
Closed
#170 Negative and boundary tests for internal/blobgen and internal/types (and internal/crypto if it survives)
Closed
#151 Remove the unused second encryption path and unused secret types; write the blob-ID hash step once
Closed
#172 config.example.yml carries a real-looking S3 key pair and internal host names
Closed
#168 Give the local index and its export copy an explicit 0600 mode
Closed
#153 Parse age_recipients when the config loads, and never print the entry in an error
Closed
#162 Open the downloaded snapshot database read-only, keep it in a private temp directory, remove it on every path
Closed
#161 Under --skip-errors a packing failure can record a file that cannot be restored
Closed
#166 config set echoes secrets and keeps a loose file mode; a storage_url carrying credentials is accepted and echoed
Closed
#160 snapshot purge deletes every local snapshot record and then purges nothing
Closed
#163 Restore: create files 0600, remove partial files, and make the blob hash check impossible to skip
Closed
#167 Reject a blob_size_limit below the largest possible chunk
Closed
#157 Prune must delete nothing when a manifest cannot be read, and a manifest must never omit a blob
Closed
#158 s3.use_ssl is documented as defaulting to true but the code defaults to plain HTTP
Closed
#150 Scope the process-wide PID lock to mutating commands
Closed
#154 Restore must keep every write inside the target directory
Closed
#148 Interrupted blob upload leaves committed chunk rows; a retry silently produces an unrestorable snapshot
Closed
#72 Fault-injection tests: prove backups survive interruption and corruption
Closed
#112 Under --json, prune deletes local index rows with no signal in either stream
Closed
#74 Documentation accuracy sweep: undocumented behavior and stale claims
Closed
#137 S3 test credentials for a private storage server remain in git history after the fixture was deleted
Closed
#124 Document and test restoring on another machine
Closed
#66 Add tests for internal/storage — the backend abstraction has zero coverage
Closed
#75 Docker builds report unknown version, and profiling output is truncated on error
Closed
#96 getTableCount errors are discarded, so a failed query reports as a count of zero
Closed
#130 Interrupted blob upload leaves a partial object the next run trusts as complete (file:// backend)
Closed
#129 S3Storer.Get/Stat return the raw SDK error, not storage.ErrNotFound
Closed
#123 Two duration parsers, and the documented --older-than example deletes almost every snapshot
Closed
#121 Rework the lint-guard shell scanner in cmd/vaultik/lintdocker_test.go (second-review findings on the milestone PR)
Closed
#131 verify --deep hashes encrypted bytes but compares to the plaintext blob ID, so it fails on healthy snapshots
Closed
#105 The Go toolchain in the release workflow is version-pinned but not hash-verified
Closed
#120 snapshot create fails on any host without the sqlite3 CLI: VACUUM shells out instead of using the driver
Closed
#70 Delete dead code and stale fixtures before tagging
Closed
#67 Docs misstate the remote layout and understate the privacy guarantee
Closed
#122 CI never runs on next or on PRs targeting it
Closed
#87 --cron flag help text misstates what it suppresses
Closed
#116 Guard scanner: shellCode silently discards the rest of a file on an unterminated here-document
Closed
#117 Guard scanner: assertLinterIsContainerised accepts a || host-linter fallback
Closed
#118 Dockerfile.lint citation names an lll finding where the PR records revive
Closed
#115 Host-lint guard scanner goes blind on a phantom heredoc and accepts a || host fallback
174 Issues created by 2 users
Opened
#116 Guard scanner: shellCode silently discards the rest of a file on an unterminated here-document
Opened
#117 Guard scanner: assertLinterIsContainerised accepts a || host-linter fallback
Opened
#118 Dockerfile.lint citation names an lll finding where the PR records revive
Opened
#119 Review toward 1.0: identify and file all next steps
Opened
#120 snapshot create fails on any host without the sqlite3 CLI: VACUUM shells out instead of using the driver
Opened
#121 Rework the lint-guard shell scanner in cmd/vaultik/lintdocker_test.go (second-review findings on the milestone PR)
Opened
#122 CI never runs on next or on PRs targeting it
Opened
#123 Two duration parsers, and the documented --older-than example deletes almost every snapshot
Opened
#124 Document and test restoring on another machine
Opened
#125 Work remaining before vaultik 1.0
Opened
#129 S3Storer.Get/Stat return the raw SDK error, not storage.ErrNotFound
Opened
#130 Interrupted blob upload leaves a partial object the next run trusts as complete (file:// backend)
Opened
#131 verify --deep hashes encrypted bytes but compares to the plaintext blob ID, so it fails on healthy snapshots
Opened
#133 CI runner mac-studio fails every job before the first step (docker network pools exhausted, disk full)
Opened
#137 S3 test credentials for a private storage server remain in git history after the fixture was deleted
Opened
#148 Interrupted blob upload leaves committed chunk rows; a retry silently produces an unrestorable snapshot
Opened
#149 Route direct-stdout command output through internal/ui
Opened
#150 Scope the process-wide PID lock to mutating commands
Opened
#151 Remove the unused second encryption path and unused secret types; write the blob-ID hash step once
Opened
#152 A stream cut right after the age header reads as valid and empty, and restore then reports success
Opened
#153 Parse age_recipients when the config loads, and never print the entry in an error
Opened
#154 Restore must keep every write inside the target directory
Opened
#155 Check blob hashes, offsets and lengths read from the destination before using them
Opened
#156 Restore and deep verify must check the decrypted database is the snapshot that was asked for
Opened
#157 Prune must delete nothing when a manifest cannot be read, and a manifest must never omit a blob
Opened
#158 s3.use_ssl is documented as defaulting to true but the code defaults to plain HTTP
Opened
#159 Ctrl-C during restore exits before cleanup, leaving decrypted data in the temp directory
Opened
#160 snapshot purge deletes every local snapshot record and then purges nothing
Opened
#161 Under --skip-errors a packing failure can record a file that cannot be restored
Opened
#162 Open the downloaded snapshot database read-only, keep it in a private temp directory, remove it on every path
Opened
#163 Restore: create files 0600, remove partial files, and make the blob hash check impossible to skip
Opened
#164 Bound what a downloaded object may expand to, and escape control characters from the destination before printing
Opened
#165 Parse the age private key once, accept every identity in it, and stop showing it typed on the command line
Opened
#166 config set echoes secrets and keeps a loose file mode; a storage_url carrying credentials is accepted and echoed
Opened
#167 Reject a blob_size_limit below the largest possible chunk
Opened
#168 Give the local index and its export copy an explicit 0600 mode
Opened
#169 Shallow snapshot verify ignores object sizes and reports more than it checked
Opened
#170 Negative and boundary tests for internal/blobgen and internal/types (and internal/crypto if it survives)
Opened
#171 Correct the security claims in the docs and comments, and write down the accepted risks
Opened
#172 config.example.yml carries a real-looking S3 key pair and internal host names
Opened
#177 Interrupted metadata export leaves a snapshot silently treated as complete but unrestorable
Opened
#204 Daemon mode after 1.0: watch for file changes so frequent backups skip the full scan
Opened
#206 Run every build, test and lint step in the Dockerfiles at nice -n 19
Opened
#207 The disk-full restore test is still skipped although the fix it waited for has landed
Opened
#208 README roadmap and TODO.md still list finished 1.0 work as remaining
Opened
#211 docker build . stamps dev, not the git tag or short commit
Opened
#213 Re-vendor the canonical files from sneak/prompts at dd4027b
Opened
#214 A backup reports success but cannot restore a changed file after the snapshot holding its current version is dropped
Opened
#215 A retention duration with unparsed characters is misread: --keep-newer-than 1.0y deletes every snapshot
Opened
#216 File ownership is never recorded, so a restore as root makes every file root:root
Opened
#217 The local index ignores its WAL and busy-timeout settings, so a read-only command can abort a running backup
Opened
#218 snapshot restore --skip-errors aborts on a missing or damaged blob instead of skipping the files that need it
Opened
#219 Restore applies directory modes and times before writing the contents, and file modes before ownership
Opened
#220 A file:// destination that is missing (an unmounted volume) is treated as present and empty
Opened
#221 The documented "restoring on another machine" steps fail at the first command
Opened
#222 With an s3://bucket/prefix URL, objects are stored as prefixblobs/... instead of prefix/blobs/...
Opened
#223 Path filters match by case-insensitive string prefix: restoring /home/u/doc also restores doc2, Doc3 and doc.txt.bak
Opened
#224 Command output breaks the README's stdout and stderr rules in three places
Opened
#225 Backup statistics are wrong in the printed summary and in the snapshots table
Opened
#226 A file rewritten within the same second as its recorded mtime is never backed up again
Opened
#227 The process lock is check-then-write, so two writers started together can both acquire it
Opened
#228 remote info counts a live snapshot's blobs as orphaned when a manifest cannot be read, and prints store names unescaped
Opened
#229 config set cannot store a string that looks like a number
Opened
#230 A hostname containing _ makes per-name retention silently do nothing
Opened
#231 A malformed snapshot database crashes restore with a Go panic instead of an error
Opened
#232 s3.part_size is documented and accepted but has no effect
Opened
#233 Doc and help sentences that are false about the current tree
Opened
#235 The internal/vaultik and internal/database tests take longer than the 20-second target
Opened
#239 How should vaultik tell an unplugged file:// volume from a new, empty destination?
Opened
#240 Two misleading messages: a credentials warning for a config with no credentials, and a doubled error prefix
Opened
#246 The pre-commit hook fails on every commit: go.mod marks a directly imported module as indirect
Opened
#251 snapshot remove --json prints its remote-cleanup warning on stdout, ahead of the document
Opened
#253 A backup without --cron panics when a snapshot has more than one path
Opened
#266 A blob upload to an rclone destination cut off by a hard kill leaves a partial object that the next backup trusts
Opened
#267 An interrupted snapshot create, restore or verify exits 0 and prints nothing
Opened
#268 The help and README do not say that --json skips the confirmation prompt of snapshot remove and prune
Opened
#269 A symlink whose target cannot be read is left out of the snapshot silently, even without --skip-errors
Opened
#270 Two snapshot create runs of one name within the same second fail on the snapshot ID
Opened
#271 Progress lines over-report on the second and later paths of a multi-path snapshot
Opened
#272 remote info shows 0 blobs and 0 B for a snapshot whose manifest cannot be read
Opened
#280 A file skipped while a backup is processed is reported as "unchanged" and its bytes as "backed up"
Opened
#281 remote nuke reports the destination store empty while a leftover .partial upload remains
Opened
#284 Under --skip-errors, a local index error while recording a directory or symlink is skipped instead of stopping the backup
Opened
#286 An interrupted --skip-errors backup keeps opening every remaining file and logs each one as failed
Opened
#288 Defect hunt on next before the release candidate
Opened
#289 A retention duration of 293 years or more deletes every snapshot
Opened
#290 prune and remote nuke exit 0 when a delete fails, and nuke still says the store is empty
Opened
#291 remote nuke leaves a snapshot database without a manifest, and non-blob files under blobs/, and says the store is empty
Opened
#292 remote nuke against a destination it cannot list deletes the local index first
Opened
#293 snapshot remove reports success for a snapshot that does not exist
Opened
#294 --cron and --quiet do not silence the status lines of prune
Opened
#295 remote info counts a file under blobs/ that is not a blob as an orphaned blob
Opened
#296 A symlinked metadata/ or blobs/ directory in a file:// destination lists as empty, and prune then deletes every blob
Opened
#297 An S3 listing the server marks truncated without a continuation token is taken as complete, so prune can delete blobs still in use
Opened
#298 With the s3.* configuration, a changed bucket or prefix is not detected, and the next backup reports success but cannot be restored
Opened
#299 A relative file:// path is accepted, so the destination moves with the working directory and a backup can exit 0 and be unrestorable
Opened
#300 A reused blob is recorded with the size of a local copy that was never stored, so snapshot verify fails on a good snapshot
Opened
#301 Ctrl-C during a backup can end with "FATAL: failed to rollback transaction" and exit 1 instead of the interrupted exit
Opened
#302 A failed or interrupted backup leaves its half-written blob in the temp directory
Opened
#303 Exclude patterns starting with **/ miss a match directly under the source path
Opened
#304 The "deleted" figure of a backup counts files that were not deleted since the previous snapshot of that name
Opened
#305 Directories and symlinks are counted as files to back up on every run, so an unchanged tree reports data backed up
Opened
#306 The index cleanup before each backup reports total rows as "remain"
Opened
#307 An invalid --keep-newer-than is checked only after the backup, which is then reported as failed and retention never runs
Opened
#308 config set with a list index on a key that does not exist writes a map, and then no command can load the config
Opened
#309 snapshot restore ignores a requested path that matches nothing, and exits 0
Opened
#310 snapshot create --skip-errors exits 0 and prints "Finished successfully." after skipping files
Opened
#311 snapshot list shows a NEW CHUNK SIZE of 0 B for a snapshot whose change is a modified file
Opened
#312 snapshot list and the snapshot purge listing print 1024-based sizes labelled MB, unlike every other command
Opened
#313 Byte-size settings: the defaults and the chunk-size minimum are 1024-based while the docs, the parser and the messages are 1000-based
Opened
#314 snapshot verify --deep --json reports "missing": 0 and "verified": 0 when it fails
Opened
#315 config edit fails when $EDITOR carries arguments
Opened
#316 Every restore by a non-root user warns that ownership was not preserved, even when it was
Opened
#317 After storage_url changes, snapshot list advises prune, which then refuses
Opened
#318 Shallow snapshot verify reports any error reading a blob's size as "Missing", and drops the error
Opened
#319 A mistyped subcommand under snapshot, config, remote, database or completion exits 0 and does nothing
Opened
#320 A config error that contains "): " loses its start, including which setting is wrong
Opened
#321 RCLONE_CONFIG is ignored, so an rclone destination configured there is not found
Opened
#322 Credentials can still be put in storage_url, inside the endpoint or an rclone remote's parameters, and are then printed
Opened
#323 restore --skip-errors --verify reports a file that failed to restore again as a checksum failure
Opened
#324 A restore of selected paths keeps decrypted blobs in the temporary cache until it ends
Opened
#325 Interrupting a restore during a slow blob read leaves the decrypted snapshot database and decrypted file data in the temp directory
Opened
#326 Under --skip-errors, one file whose chunk is in no blob aborts the whole restore before anything is restored
Opened
#327 Restore and verify error lines repeat a prefix, and deep verify names S3 for any destination
Opened
#332 A file:// destination whose metadata/ or blobs/ is a symlink to a missing target lists as empty, so prune deletes every blob
Opened
#334 Record the defect-hunt fixes in TODO.md Completed Steps
Opened
#337 snapshot purge exits 0 and still deletes the remote metadata when it cannot remove a snapshot from the local index
Opened
#338 Exclude patterns /**/x and a/**/b do not match as in gitignore
Opened
#340 snapshot remove exits 0 when deleting a destination-only snapshot's metadata fails, and its README promise of a prune command is false on several paths
Opened
#345 internal/database and the blob upload step still end the process on a rollback after Ctrl-C, a failed close of query results, or a failed temp file cleanup
Opened
#351 The Docker build hangs for as long as a stalled connection to the Go module proxy lasts
Opened
#356 Interrupting a stalled read of the snapshot database leaves decrypted data in the temp directory
Opened
#360 snapshot list's UNCOMPRESSED SIZE of an older snapshot changes after a later backup
Opened
#382 snapshot create --skip-errors --prune deletes the previous snapshot, the only one holding the skipped files
Opened
#387 A prune on one host deletes blobs of another host's backup in progress; that backup reports success and cannot be restored
Opened
#388 A backup of a source directory left empty by an unmounted volume deletes every older snapshot under --prune
Opened
#389 The documented PUT-only source host fails every backup after its first: the destination marker check needs HEAD permission on S3
Opened
#390 prune never removes the encrypted database a backup left on the destination when its manifest upload failed
Opened
#395 A backup while a volume mounted below a configured path is not mounted deletes the older snapshots holding its files under --prune
Opened
#397 Rewrite README.md and the other prose documents sentence by sentence, on Fable
Opened
#398 snapshot restore mentions VAULTIK_AGE_SECRET_KEY but snapshot verify does not
Opened
#400 snapshot remove refuses the remote key that snapshot list shows for a remote-only snapshot
Opened
#401 Missing-key errors for restore and deep verify name only VAULTIK_AGE_SECRET_KEY
Opened
#402 Terminal log output ignores NO_COLOR
Opened
#403 Two code comments describe behaviour that no longer exists
Opened
#408 config init writes a config whose first backup on Linux stops on /Applications
Opened
#409 vaultik version labels the commit date "build date"
Opened
#410 snapshot verify and remote info print status lines without internal/ui
Opened
#411 Format Markdown with prettier and move documents out of the root, as REPO_POLICIES.md requires
Opened
#416 internal/vaultik tests take 36-56 seconds and pass the 90-second timeout when the host is busy
Opened
#418 Acceptance test: use vaultik end to end as a user before a release candidate
Opened
#419 A misspelled or unknown config key is accepted silently
Opened
#420 A backup that fails before writing anything binds the local index to the configured destination
Opened
#421 The default config backs up vaultik's own live local index and lock file
Opened
#422 A binary from go install reports version dev although it was built from the tagged release
Opened
#423 cosmetic: a cancelled confirmation prompt exits 0
Opened
#424 cosmetic: a corrupted blob's raw bytes are printed in every error, once per file under --skip-errors
Opened
#425 proposal: remote info reports snapshot blobs missing from the destination
Opened
#426 cosmetic: restore and verify of an unknown snapshot ID report "object not found" on a hashed path
Opened
#427 cosmetic: the wrong-private-key error says "no identity matched any of the recipients" with a doubled prefix
Opened
#428 cosmetic: config set storage_url to s3:// is refused until the S3 keys are set
Opened
#429 cosmetic: a non-numeric chunk_size or blob_size_limit fails with an error that names neither the setting nor the value
Opened
#430 cosmetic: index_path accepts a relative path
Opened
#431 cosmetic: config init --help names the config directory without its vaultik component
Opened
#432 proposal: a --version flag
Opened
#433 cosmetic: a failed command prints its error twice on stderr
Opened
#434 cosmetic: Ctrl-C during a backup logs an upload error before the interrupted message
Opened
#435 proposal: list the files in a snapshot without restoring it
Opened
#436 proposal: a dry run for snapshot create that shows what the excludes keep and drop
Opened
#437 proposal: retention by count and by day, week and month
Opened
#438 proposal: restore a path into a directory without recreating its absolute path
Opened
#439 proposal: a distinct exit status for a backup or restore that finished but skipped files
Opened
#440 TODO: remove work logs; keep only open work
Opened
#455 An exclude pattern that does not compile is skipped during the backup instead of refused when the config loads