The no-effort rule was keyed on every claude-haiku id, so Haiku 5.5 ran at the engine default instead of its phase effort (mechanical: low). The guard now names Haiku 4.x; status-reporter pins effort: low.
CLI 2.1.306 prints an advisory plugin-directory section (ICON_MISSING,
blocking: false) that the whole-output `warning` grep counted as a
failure. The verdict is now the rc of `claude plugin validate --strict`
(per-file warnings → 1, directory advisory never), through one helper
shared with a positive control on two scratch manifests.
New best-tier phase audit (xhigh) carries both merge gates; their opus frontmatter stays the outage floor, since inTier never writes below a tier head. The census floors best-tier rows on the second alias, and the citers now name the audit row.
The probe trusted the plan's Supabase url, followed redirects for free,
crashed on deep JSON and stopped at the first unconfirmed map candidate.
Rules now scope hydrated state to the export chunk, silence dev configs,
UI-suffixed data attributes, GraphQL types and metadata keys, and catch
Go raw strings, triple-quoted f-strings, sql.raw, empty select() and
multi-line hidden inputs; text output is sanitised; per-variant locks.
After the report the skill proposes fixes (plan phase), stops for the
user's acceptance, applies accepted ones via bugfixer and the
verify-secure gates, then re-scans. scan.py gains --assert-absent
(contract oracle, positive control) and --baseline (no new hit).
probe.py observes a running target the user owns and names, driven by a
plan built from the static findings. Five probes: anon-read, cross-user,
no-auth route, CORS reflection, capped rate-limit burst. GET/HEAD/OPTIONS
only, --authorized gate, secrets from the environment, same-host redirect
only, bounded request budget. SKILL STEP 5, checks.md reference, docs.
Suite guardrails-probe.test.sh (16 checks) against a localhost fixture.
Eval iteration 1: the callout 'edge' variant picked up the diagram connector styles, and all three skill runs hand-wrote a template filler. fill_template.py takes values.json with @fragment files, escapes </ in the quiz; suite 21 checks.
A1 drop unused `import gzip`; B1/B2 file-level shellcheck directives
placed before `set -u`; B3 `cd "$REPO" || exit 1`; B4 hook regex split
into `pattern+=` lines (sha256 of $pattern unchanged); B11
`_build_parser`/`_dispatch` pulled out of google_seo `_cli`; B20
detect-plugins header names update-all.sh; C2 prune-memory deterministic
suite wired into SUITES (literal path); C3 `_refuse_outside_allowlist`
moved up in deploy-commit.
No behavior change except B3 (a failed cd now exits 1) and C2 (one more
suite). B5/B7/B9/B18/B19 land in the refactor commit.
Kill switch fails closed: a present-but-unreadable override at first load
or a non-boolean enabled turns the router off (logged); /route reload says
when the previous config was kept. Prompt rules with nested quantifiers
are dropped; a config path that is not a regular file reads as failed;
caught errors log kind + code only (errWord); rejected keys are masked in
logs (keyWord), alias keys match KEY_NAME, prototype names are never keys;
offers capped at 256; a models alias without a fallback list is ranked at
the end of the default chain; canonical matches a table id at a '-'
boundary only; logCall/noteMain cannot unroute a step; route-tool failure
text no longer claims nothing was routed.
Repo side: mods suite probes the 'claude plugin test' usage line, doctor
bounds plugin list to 60 s, update-all skips @skills-dir mods.
Kit 431/431, full make test green.
The text regex becomes a prefilter. The command is split into simple
commands by an awk tokenizer (quotes, escapes, ; && || | ( ) backticks,
heredoc bodies dropped) and a push is only a simple command whose argv[0]
is git (or */git) with push or send-pack as its first non-option word.
grep "git push", git log -S, git config --get push.default, git stash
push, heredocs and commit messages quoting a push are no longer denied.
bash|sh|zsh -c scripts are re-checked (depth 3); wrappers (sudo, env,
eval, xargs, timeout, git subtree...), assignments whose value reads as
a push, a missing awk or > 2000 simple commands keep the text verdict.
Prefilter also matches "git" push (quote characters stripped).
Suite 98 -> 124 checks.
routing.json (tracked, reached through the plugin directory) is now the
single source of the phase table and of every skill/agent row, plus the
decisions: confirmed rows/phases, changed rows (from/to) and projects
exceptions keyed by a normalized git remote (credentials never stored, no
machine paths). First use of a rowed typed skill, a rowed agent spawn or a
main-loop phase opens the engine's dialog (Later / Keep / two alternative
phases; Other = a phase name); a change asks Everywhere or This project
only. One dialog at a time, never in headless, never inside an agent,
never written by the model: only a dialog answer or /route ask writes,
serialized, size-capped, never creating the file. Layers: routing.json <
~/.claude/model-router.json; the project tree is never read. /route
pending, /route ask on|off. Census reads rows and phases from the file and
tolerates a user-changed row (WARN). Kit suite 86 → 190 tests.
Contract .claude/tasks/contracts/2026-10-10-model-router-w3a-confirm-1201.md,
plan r4: 3 lenses + 1 confirmation, feater + 4 rounds, GATE 0 MET,
verifier CONFORME then re-verify after security, security BLOCK(1) fixed
then PASS. Live: T2 dialogs answered by the user from the hot-loaded mod.
The 15 Skill(effort-*) citers now call mcp__model-router__route per phase
(orchestrate at a dispatch span, reflect/plan for the skill's own level,
apply at the bookkeeping tail, escalate at the verify-secure caps); built-in
judgment dispatches carry an explicit effort= param. lib/effort-shift.md is
the route doctrine, lib/model-gate.md the mod rule (route answer = witness,
/route on as remedy). Deleted: skills/effort-*, lib/effort-pins.txt/.sh,
lib/model-check.sh, their tests, the installers' re-apply blocks. The mod
drops its Skill(effort-*) bridge. The tracked model:/effort: frontmatter
stays as the off-state floor, census-locked equal to the rows
(lib/tests/effort-routing.test.sh rewritten, 140 checks; analyzer → xhigh).
Contract .claude/tasks/contracts/2026-10-10-model-router-w2b-1045.md, plan
r4 § W2-B: GATE 0 MET, verifier ECARTS(7) then CONFORME 10/10, security
PASS, full make test green (design-tool-gate env red only).
Tracked relative symlink skills/model-router -> ../mods/model-router: Claude
Code loads the mod in place as model-router@skills-dir wherever link.sh
links ~/.claude/skills (no CLAUDE_CODE_PLUGIN_DIRS: absolute paths in the
tracked settings.json). Engine-laid mods/*/tsconfig.json gitignored.
lib/tests/mods.test.sh: manifest name, link target, claude plugin validate
and test per mod, capability-probed, time-bounded, SKIP with reason.
doctor.sh: fail-soft Mods section (link by -ef, one guarded plugin list).
CLAUDE.md: mods/ section (loading, per-machine enabled:false switch,
dev-copy shadowing, tests).
Run D2 of manual-push mode (BDR-114).
- push-guard sources lib/gitflow.sh once (absolute path) and reads each
candidate dir through gitflow_push_mode; a missing lib denies.
- Dir tokens are extracted as whole shell words: a fully quoted token
(inner apostrophe allowed) is resolved, a backslash-escaped space is
unescaped deterministically, a token mixing quoted and unquoted parts
is refused (fail closed) instead of resolving to its parent.
- A payload jq cannot parse is scanned as raw text with its JSON escapes
folded; a push-looking one gets the static deny through the trap.
- The 20-token cap runs before any per-token classification (a flood of
20 000 tokens is refused in 0.13 s; T58 locks it under 5 s).
- `case "$mode"` has a deny default; missing core tools warn and allow.
- T42 compares the deny list against main (the last release) instead of
HEAD; literal-true, mixed-token, broken-payload, lib-missing and
banner-on-bad-value cases added (98 checks).
- session-start banner reads the mode through the verb and shows
`push : manual (autopush bad)` on an unparseable value.
- tour hints quote "<abs project>".
Run D1 of manual-push mode (BDR-114). `git config --bool --default true
gitflow.autopush` only covered a MISSING key: an unparseable value made
git die with empty output, the `= false` test failed, and every push ran
again. A typo on a work machine silently re-enabled the pushes it was
meant to stop.
- lib/gitflow.sh: `_gitflow_push_off` reads the mode through the lib
verb (`push-mode`); anything but `auto` is push-off, and the verb's
stderr line names an invalid value during start/finish.
- Emitted post-commit/post-merge hooks (POSIX sh, standalone): push only
when the key reads `true` or is unset; `false` exits quietly; any
other result prints one stderr line ("NOT pushed, treated as manual
push mode") and exits 0. Mirrors gitflow_push_mode.
- .githooks/ and githooks/ regenerated files-only through `emit-hook`
(no config read or write; .git/config hash unchanged).
- hooks/unpushed-guard.sh: mode from the lib verb (absolute lib path
resolved before any cd, no temp file); anything but auto is manual;
the SessionStart line names an invalid or unreadable value.
- Tests: gitflow-test T18q block (invalid → start, hook and finish push
nothing and say so; `true` → the hook pushes; emitted hook is
POSIX-clean), unpushed-guard T14 rewritten.
Hardening after the security gate on a2ac018 (3 MEDIUM, all closed and
re-measured):
- dir tokens are deduplicated and capped: more than 20 distinct cd/-C
targets in one command denies before any git fork (20000 tokens: 0.15 s
against the 10 s hook timeout that used to turn a flood into an allow)
- a git or cd failure while reading gitflow.autopush denies instead of
reading as auto (git absent, usage error, unenterable dir); the key
being unset is the only "auto" answer; the decision is recorded only
after one candidate was evaluated cleanly, else the EXIT trap denies
- cd/pushd/-C targets that follow a quote or backtick (bash -c '…') are
extracted; quote characters are excluded from unquoted tokens
User decision (contract, gated): both fail-closed cases also fire in
auto mode on such pathological commands; silence in auto mode holds for
every ordinary push. Header limits list the residual misses (quotes or
backslashes inside a token, cumulative relative cd, unparseable payload)
backed by the soft_deny rule. Tests: 71 checks (T48–T50b added).
Run B of manual-push mode (BDR-111). With `gitflow.autopush false`
nothing stops Claude from typing `git push` itself: the `ask` tier is
inert under auto mode. This adds the mechanical block the user chose.
- hooks/push-guard.sh (PreToolUse, matcher Bash|Monitor, timeout 10):
detects a push in the command text (strict, quote-stripped loose and
alias patterns; backslash-newline folded in bash, BSD sed/grep only),
reads gitflow.autopush in the payload cwd and in every literal -C/cd
dir the command names (global config counts outside a repo), denies
with the documented JSON form and a reason that tells the user to run
the command with `!`. Unparseable value = manual (fail closed); once
a push is detected an EXIT trap emits a static deny on any internal
error. jq missing = one stderr warning, allow (sibling-hook policy).
- settings.json: hook wiring; 18 deny entries closing the write forms
of the human-only toggle (any `git … config` spelling, section
removal/rename, `-c`, config env overrides, direct edits of git config
files); one soft_deny on pushing in manual mode in any form, with no
per-turn clearance; the routing-around rule names hook refusals.
- hooks/session-start.sh: `🔒 push : manual (autopush=false) — ! git push`
banner line when the key reads false (padding in bytes).
- lib/tests/push-guard.test.sh: 61 checks (push forms, over-blocks,
invalid value, global key, fail-closed trap, no-jq, wiring, banner).
Known limits are listed in the hook header; the soft_deny rule is the
backstop. Run C (skills that push on their own) and run D (fail-closed
readers everywhere) follow. Do not enable manual mode at work before C.
`gitflow.autopush false` (human-set git config) now means "nothing is
pushed" end to end, not only in the post-commit/post-merge hooks:
- lib/gitflow.sh: `_gitflow_push_off` is the single reader of
GITFLOW_NO_PUSH / gitflow.autopush for the lib's push sites; `start`
and `finish` stop pushing in manual mode. `gitflow_delete` checks out
the base that contains the branch and drops a lagging upstream before
`git branch -d` (LRN-161: `-d` judges against the upstream when set).
Skipped remote deletes say `left in place`; `_gitflow_sync_base`
replaces the silent `pull --ff-only || true` and warns when a base is
behind origin and cannot fast-forward.
- hooks/unpushed-guard.sh: manual mode is silent at Stop and gives one
`ℹ manual push mode:` line at SessionStart counting every local
branch; an unparseable value is named and treated as auto.
- CLAUDE.global.md: manual-push mode doctrine, "ahead = defect" scoped
to auto mode.
- Tests: gitflow-test T18m block (T18m0, T18i-T18o, 7 cases),
unpushed-guard T10-T16.
Follow-ups (TODO.md): run B push-guard hook + settings deny widening +
banner; run C skills that push on their own (/close STEP 5C, …).
Do not enable manual mode on the work machine before B and C land.
The suite and seven libs assumed a GNU userland: `cmd | grep -q` under
pipefail (grep exits at the first match, the producer takes SIGPIPE,
rc 141 → 15 false "merged into" FAILs in gitflow-test), `sed -i` with no
suffix, `wc -l` padding compared as a string, `stat -c`, `touch -d`,
`realpath -m` (gstack-links refusal never fired), bare `timeout` off the
sanitized PATH (design gate READY BUT UNVERIFIED), `grep -oP` (update-all
emptied the plugin list). Thirteen suites were red on this machine.
Portable forms on the native userland of both OS: producer captured out
of the pipeline, `sed -i.bak` in tests and a temp-sibling `sed_profile`
on the user dotfile, `tr -d ' '`, python3 perms, `touch -t`, a
`realpath -m` emulation that refuses `..`, perl `alarm` for the 15 s
bound, `sed -n` token extraction. Regression tests: gstack-links T4b,
doctrine-citers Alphabet/Alpha flip, profile-set-managed T18 (failing CLI
no longer aborts `set`), new portability-census suite over the tracked
shell files. Linux run deferred (see TODO).