The slug-migration fallback (used when there is no live symlink to read
the previous slug from) recomputed the local-* candidate by lowercasing
the basename only, while _agents_repo_slug sanitizes it. The two formulas
had drifted, so the fallback silently found nothing for any project
directory whose basename needed sanitizing.
Extract the formula into a single private helper,
_agents_repo_local_slug, and have both _agents_repo_slug's no-remote
branch and agents-vault's migration fallback call it, so there is one
place left to drift.
Also drop two dead lines the review flagged: an unused local,
and an unreachable mkdir -p (path dirname ...) — slugs never contain a
path separator, so dirname always resolves to a directory that already
exists by that point.
Widen migration test coverage: the current-entry-present-but-empty case,
a remote URL rewrite, a remote removal, and a dirty-basename fallback
test that fails without the sanitization fix and passes with it.
Adding a remote to a previously remote-less project changes its slug. Left
unhandled, the link step repinned the live memory symlink to a fresh empty
entry and orphaned the real memory.
The previous slug is read from the live symlink target rather than guessed,
which covers a remote being added, rewritten, or removed. When both the old
and new entries hold content the migration is ambiguous, so nothing moves
and the user is directed to --adopt.
The guard around the symlink step only linked when the live Claude
project directory already existed, which is exactly backwards for the
clone-onto-a-new-machine restore case: a freshly cloned vault entry
would be silently left unlinked and a starting agent would write fresh,
history-less memory instead. _agents_repo_ensure_symlink already makes
its own parent directories and is idempotent, so nothing depended on
the guard; it is removed and the link is now attempted unconditionally.
Also stop swallowing a refused or failed link as success: the helper's
exit status is now checked, and agents-vault reports its own error and
exits 1 instead of silently continuing with no link in place.
Smaller fixes from the same review pass:
- check the exit status of _agents_repo_install_tools and the
core.hooksPath git config write, instead of discarding both
- give the vmem mkdir failure a stderr message like every other fatal
in the function
- guard hostname with type -q and add it to DEPENDENCIES
- .version creation now sets changed, so --link (which skips the
commit step) reports it in --quiet mode
- reword --link's help/doc text: it still scaffolds the vault and
links memory, it only skips the final commit
- drop the unused c_dim color variable
- move the __fish_agent_vault_dir / __fish_agent_vault_autopush
documentation below Opinionated Components so its NOTE: callout
(now flush-left so it actually renders as a Starlight Aside, per
review) doesn't become the first Note aside in the page and steal
the existing test's assertions about the original 4-bullet one
Adds two tests: pre-seeded vault entry with no live directory at all
(the restore path the guard was breaking), and a forced link failure
asserting agents-vault now exits 1 instead of 0.
Creates the vault repo on demand, reusing the AGENTS version bumper and
hook shims, then links the current project's live memory directory into
its slug-keyed entry and commits.
Because the live directory becomes a symlink into the vault, backup and
restore are the same operation: a cloned vault relinks itself on the next
run in each project, with no manifest and no batch restore step.
Also fixes _agents_repo_install_tools' progress messages, which hardcoded
the literal "AGENTS/.agents-tools/" even for callers writing elsewhere:
they now name repo_dir's own basename, so agents-vault reports its own
directory instead of a false AGENTS/ path.
Renames _agents_init_install_tools to _agents_repo_install_tools now that
the vault shares it, collapses the two duplicated root-symlink blocks into
one loop, and routes the auto-commit through _agents_repo_sync so a failed
rebase can no longer be committed as conflict markers.
_agents_repo_sync fell off the end of its final if-block when git commit
failed, e.g. a pre-commit or commit-msg hook rejecting it (this repo runs
ggshield and Git-LFS hooks). fish's if construct sets status 0 when the
condition is false and there is no else branch, so a rejected commit was
being reported as success rather than as the documented "not a git
repository" exit 1 it was assumed to fall through to.
Add an explicit else branch that emits a stderr diagnostic and returns 1,
and widen the EXIT STATUS/DESCRIPTION docs to cover this path under the
existing code 1 rather than adding a fourth code, since later tasks
already consume the 0/1/2 contract.
agents-init currently swallows a failed rebase and then stages and commits
whatever is in the tree, which records conflict markers under a routine
message. No AGENTS repo has a remote today so the pull never runs, but the
vault gives these repos remotes and arms it.
The shared helper aborts the rebase, commits nothing, and returns 2. It
also redirects git's own stdout during the pull/abort: git prints
"CONFLICT (content): ..." to stdout, not stderr, so without this the
message would leak into the helper's own stdout instead of staying
diagnostic-only.
The conflict fixture commits "ours" locally before diverging, since an
uncommitted worktree change has nothing for --autostash's rebase step to
replay -- it fast-forwards cleanly and only the stash pop would conflict.
Enforces the rails the vault depends on: only directories are linked
(agent editing tools refuse to write through a symlinked file), a missing
target is refused rather than turned into a dangling link, and adopting a
populated live directory copies without clobbering.
Also fixes slug sanitization in _agents_repo_slug to apply the same
[^a-z0-9._-] → - mapping to the fallback (no-remote) branch, ensuring
local project slugs are filesystem-safe and won't leak special chars like
spaces or exclamation marks.
Keys a project by its remote rather than its path so the key survives a
machine change or a directory rename. Falls back to a path-derived
local-* key when no remote exists.
Adds a hermetic test harness that builds throwaway repos under mktemp.
The heading scan matches a keyword contained in a heading, so a plural
could never reach a singular heading: `customization` found "7.
CUSTOMIZATION" and `customizations` found nothing at all. The index
lookup was exact-match, so it could not cover the gap either without an
alias per word.
Both now try the keyword as typed first, then its singular/plural
forms. Each variant is tried against every heading before the next one
is considered, so a loose plural cannot beat an exact hit further down
the document.
config-help pipes the manual through bat, which dims the backticks and
leaves the span content the same colour as the prose around it -- so a
delimiter carried no information and every span cost the reader two
literal characters. 1236 of them across the document.
Each span is now rendered bold and the delimiters dropped, in every
branch of the viewer chain:
- ov + bat, and ov alone, style the spans (bat's output wraps each
backtick in its own SGR sequence; raw Markdown is matched directly)
- bat alone flattens them on the way in instead, because bat escapes
any SGR sequence handed to it as input
- less and cat style them, less gaining -R to render the result
- man -l needs nothing; pandoc consumed the backticks at build time
Both substitutions are line-preserving, so the tail-slice that opens
the pager on a requested section still lands on it.
CONTRIBUTING states doc-headers are written as plain text -- the header
is read as-is by config-help, by funcsave, and by anyone opening the
file, and docs/codespans.py adds the site's inline code spans at render
time. 22 files had drifted from that, carrying 41 hand-written spans
that reached config-help and the man page as literal backtick
characters inside an otherwise verbatim block.
The one span whose content ended in a space is requoted rather than
dropped, so "read> " keeps reading as a prompt string.
Add the established c_head/c_cmd/c_flag/c_dim/c_arg color scheme to
--help (or usage-on-error) output in play-media and 13 other functions
that lacked it or used an ad hoc scheme: bkg, detach, replay, p, y,
spark, wake-lock, open-url, repo-open, dng2avif, dockup, fish-deps
(__fish_deps_help), and scrub.
Also tweak the standard itself:
- c_cmd now uses plain `set_color --bold` instead of `--bold white`,
so the command name adapts to the terminal's foreground instead of
forcing white text that washes out on light-background themes.
Applied across all functions already using the pattern.
- jobrunner's reset variable renamed from c_rst to c_reset to match
the naming used everywhere else.
A missing trailing backslash after the Thumbs.db:encryptable pattern
terminated the `set -l aggressive_patterns` array early. The AI-tool
patterns (.gemini*, .claude*, .antigravity*, .remember*) were then
executed as a bogus command instead of being appended to the array,
so scrub -a never actually purged them, and every scrub invocation
printed a spurious "Unknown command" error.
Fuzzy-filters audio/video files under the current directory via fd,
previews them with thumbnails pulled from the freedesktop thumbnail
cache (or ffprobe metadata as a fallback), and plays the selection
through the best available player: --player flag, $play_media_player,
xdg-mime default, then a short mpv/vlc fallback list. Adds mpv/vlc as
optional fish-deps entries.
The @@ picker only listed files with no preview. It now lists both
files and directories via fd (matching _fzf_search_directory), and
shows a bat-highlighted or image-rendered preview through
_fzf_preview_file. Image previews use a kitty-graphics-protocol,
chafa, viu, timg fallback chain via the new _fzf_preview_image
helper, benefiting the Ctrl+F directory search and git-status
pickers as well since they share the same preview helper.
Binding the raw @@ chord made a lone @ an ambiguous prefix, so fish
(with fish_sequence_key_delay_ms unset) held every typed @ indefinitely
until a disambiguating keystroke arrived, breaking things like
`ssh user@host`. Binding the single @ key instead and checking whether
the current token is already a bare @ removes the ambiguity entirely:
plain @ always self-inserts instantly, and a second consecutive @
triggers the picker in place, with no forced delimiter before you can
keep typing. Vi's normal/visual modes are left unbound, matching their
existing (no-op) @ behavior.
The wrapped description for the Enter key was indented one column past
every other description in the Navigation block. Drop the extra leading
space so it lines up at column 17 with the rest.
Bundled fix wave for six findings from the sub-category-granularity
whole-branch review:
- config-settings: reset in_subcat on Tab/Shift-Tab so up/down keep
routing to cur_row instead of freezing on the Sponge/Paths pages
after a drill-down
- verify-manual: _parsed_components() now delegates to
generate_component_registry.collect_components() instead of a
lossy dict.update() merge, closing a taxonomy-check blind spot on
identity collisions
- verify-manual: add test_committed_registry_matches_headers to
catch the committed registry drifting from current # COMPONENT
headers
- build-docs workflow: trigger on conf.d/** and config.fish edits,
not just functions/**
- __config_settings_draw_subcat: show (Universal)/(Session) in the
drill-down title so the persistence scope is visible before toggling
- __config_settings_draw / config-settings: mention the Enter
sub-category drill-down in the on-screen hint and --help text
- generate_component_registry: send the "wrote ..." progress line to
stderr so it no longer corrupts --concat's stdout output
Two bugs invisible to single-frame rendering, only reachable via the
interactive event loop:
- right/l and left/h resolved the sub-category variable name with
"$toggle_vars[(math ...)]"_(...) inside a quoted string -- fish
cannot expand a command-substitution index there ("Invalid index
value"), so the set never ran and varname silently kept the parent
category variable. Every arrow-key press on a sub-category row
toggled the parent category instead. Fixed by hoisting the category
variable into a plain local first, the same technique the down/j
case already used for its page index.
- __cs_dispatch_draw left panel_h fixed at 16 regardless of what it
actually drew, but the new sub-category page is n+7 lines (9-13,
always < 16). Every redraw/cleanup erase computed its height from
the stale constant, erasing too many rows and corrupting whatever
was above the panel. Fixed by having the dispatcher record the
actual printed height into panel_h after every draw, including
resetting it back to 16 on the value pages even when in_subcat is
stale from a Tab away without an intervening Escape.
Adds __config_settings_draw_subcat.fish, rendering a category's own
toggle plus its sub-category rows, sized dynamically from
__config_settings_subcats instead of the fixed 6-row category layout.
Wires the real per-category row count into config-settings.fish's
up/down handling (replacing Task 17's stub) and adds the
hyphen-to-underscore sub-category variable resolution to the
left/right toggle handlers, mirroring the translation the guard
function already applies.
Two box-drawing bugs surfaced while adapting the design doc's draft to
the real static sub-category data and were fixed rather than
transcribed: the title-dashes offset was off by 5, and several real
sub-category labels/descriptions (e.g. "Notifications", 13 chars;
several descriptions past 30 chars) exceed the narrower width tiers'
fixed field widths, so both fields are now defensively truncated
before padding to keep every row exactly iw+2 wide regardless of
content length.
Adds multi-site `# COMPONENT` headers and converts every guard call site in
conf.d/tricks.fish, conf.d/abbr.fish, functions/smart_exit.fish,
conf.d/yay-wrapper.fish, conf.d/paru-wrapper.fish, conf.d/wakatime.fish,
conf.d/first_run.fish, and config.fish (9 sites across 3 categories) to the
self-identifying `(status current-function|basename) <site>` calling
convention. Regenerates conf.d/__fish_config_op_registry.fish.
__fish_config_op_enabled built the sub-category override variable name by
concatenating the tag's slug verbatim, e.g. __fish_config_op_integrations_window-mgmt.
Fish variable names cannot contain hyphens, so any hyphenated sub-category
slug (14 of the 24 in the taxonomy) silently could never be overridden --
an unset such variable safely fell through via set -q, but attempting to
set it always errored with "invalid variable name", masking the defect
since no prior smoke test exercised an explicit sub-category-level toggle.
The registry tag itself (as authored in # COMPONENT headers and the
taxonomy docs) stays hyphenated; only the derived fish variable name is
translated.
Reworks fish-deps' dependency classification from three tiers to five:
Required, Recommended, Optional, Terminal Emulators, and Integrations.
- Add an Optional tier for single-purpose wrapper conveniences that only
matter if you already use that specific tool (btop, dust, duf,
prettyping, go, lazygit, lazydocker, docker, yt-dlp, screen). Skipped by
`fish-deps install`/`sync` by default.
- Split kitty/wezterm into their own Terminal Emulators tier, since only
the one matching $TERM is ever relevant to a given user.
- Add --optional, --terminals, and --all flags to `install`/`sync` to
opt back into the skipped tiers, with a summary of how many were
skipped and which flag restores them.
- Fix `_fish_deps_status` marking missing Integrations as critical (red)
the same as Required — only Required is red now; Recommended stays
yellow, everything else renders as a neutral dim note.
Also fixes two bugs surfaced during dependency testing on a fresh
install:
- `fish_right_prompt` called `docker context show` on every prompt with
no `type -q docker` guard, unlike every other optional integration in
this config — on a system without docker this printed a visible
"Unknown command: docker" block on every single prompt render. Added
the missing guard, and hardened the `ld` wrapper with explicit
docker/lazydocker presence checks.
- `ov`'s catalog entry offered `cargo install ov` as its preferred
install method, but crates.io's `ov` is an unrelated crate, not the
noborus/ov pager — cargo would silently install the wrong package.
Removed the cargo path; `ov` now prefers `go install
github.com/noborus/ov@latest` (ahead of the system PM) when go is
available, since not all distros package `ov` in their base repos,
and falls back to the system PM otherwise.
The EXIT STATUS block used a single space before the ">0" row's
description, which falls below the doc pipeline's 2-space column
threshold and silently degrades the rendered site page from a table
(like rm's) to a plain indented block. Pad it to match. Also strip
trailing whitespace from two DESCRIPTION lines and add --wraps='mv'
so fish inherits real mv's completions, matching the bash/top wrapper
convention.
Adds jobrunner (short alias jr) for running, listing, inspecting,
re-attaching to, and terminating detached background tasks. Unlike bkg
and detach, which discard output, a jobrunner job keeps a live terminal
that survives closing the shell and can be restored later with attach.
Includes the __jobrunner_sessions parser shared by the function and its
completions, dynamic job-name completions, and screen in the fish-deps
catalog.
The agy tool uses -c and --continue natively. It does not recognize -r
or --resume. This commit allows -r and --resume to be used by
translating them to -c and --continue respectively.
Session logging is the one opinionated category that writes a persistent
record of terminal output to disk, and those logs can contain secrets.
Silent-by-default is the wrong posture for it, so C5 now defaults to off:
__fish_config_op_logging must be set to an explicit truthy value to enable
capture, and the master switch cannot enable it on its own.
Implemented as a single special case in __fish_config_op_enabled, so all
five capture components, the sentinel file, the paru/yay wrappers, the
Kitty watcher, and the config-settings TUI pick it up with no other code
changes. The existing startup sync in conf.d/logging-events.fish
reconciles the sentinel and removes the generated wrappers on the first
shell after upgrading.
BREAKING CHANGE: users relying on the previous default must now run
`set -U __fish_config_op_logging on`. Existing logs in ~/.terminal_history
are left untouched.
rootiest