feat(doctor): ask agy which model it runs; validate permissions.allow

.claude-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "antigravity",
-  "version": "0.22.4",
+  "version": "0.22.5",
   "description": "Run the Antigravity CLI (Gemini) as a collaborating AI inside Claude Code, with intelligent model routing across the SDLC. Claude conducts — requirements, architecture, the hard 20%, verification, review — and routes deterministic, high-volume work (scaffolding, tests, first-pass review, migrations, web/Vertex AI Search) to Antigravity. Hybrid agentic engineering; lower token cost as a financial lever.",
   "author": {
     "name": "linyuting"

CHANGELOG.md

@@ -3,6 +3,121 @@
 All notable changes to **Antigravity for Claude Code**. Format loosely follows
 [Keep a Changelog](https://keepachangelog.com/); versions are in `.claude-plugin/plugin.json`.
 
+## 0.22.5
+- **`doctor` now asks agy which model it will run, instead of inferring it from a version
+  string.** 0.22.4 added a warning for agy below 1.1.10, where `--model` was ignored in
+  headless `-p`. That warning is the best a version comparison can do, and a version
+  comparison is a proxy: it is right about the releases we know about and silent about
+  every other way the flag can fail to land. agy 1.1.11 answers the read-only slash
+  commands in print mode without starting an agent turn, so `doctor` requests a tier model
+  with `-p /model` and reports which one comes back — `usage.total_tokens: 0`, no quota
+  spent, no conversation left behind. It reads the tab-separated reply's slug and matches
+  it against a tier configured as a display name, the same either-direction comparison
+  `agy models` needed in 0.20.x.
+  **Gated at 1.1.11 and tested as a fact, not as prose.** Below that version the slash
+  command is not recognised, falls through as literal prompt text, and the model answers as
+  though it had run — so probing there would spend a real turn *and* then trust the answer
+  it invented. The stub agy in the suite logs every invocation, so "never probes below
+  1.1.11" is asserted against the log. An empty answer draws no conclusion in either
+  direction: an older build than the version claims, a hang, or a plan that refuses the
+  probe is not evidence that routing is broken.
+- **`doctor` validates `permissions.allow` entries.** The plugin recommends such a rule in
+  eight places as the *narrow* alternative to `--yolo`, and the recommendation ships a
+  placeholder — `write_file(<dir>)`. A rule agy cannot parse announces itself in neither
+  direction, and which way it fails depends on the version: from **1.1.11** it matches
+  nothing, so the grant is absent and the write is soft-denied (exit 15) with the rule
+  sitting visibly in the file; **before 1.1.11** an entry that tokenized to zero command
+  words matched **every** command and silently auto-approved anything the agent ran —
+  broader than the `--yolo` it was chosen instead of. `doctor` flags the entry, names the
+  reason, and scopes the consequence **twice**: to your agy version, and to the reason.
+  Every class means the grant is absent; only a `command(...)` rule naming no command also
+  carries the match-everything history, so that sentence is printed only when such an entry
+  is actually present. A mistyped `write_file()` never had it, and putting a security claim
+  in front of someone holding one would be worse than saying less. (Both reviewers caught
+  the first cut attaching it to every finding.)
+  The zero-command-word test follows upstream's own examples (`command(time)` — a shell
+  reserved word that prefixes a command without being one — a comment-only entry, and an
+  empty compound `()`), plus the unsubstituted `<...>` placeholder, which is ours — matched
+  as that *shape*, not as a bare angle bracket, so a rule carrying a literal redirect
+  (`command(echo hi > /tmp/f)`) is not misread as a template nobody filled in. Rules it
+  cannot judge are left alone: unbalanced quotes are agy's parser's business, and
+  `write_file(...)` is a different matcher from `command(...)`. A false positive here sends
+  someone to edit a rule that was always fine, so the well-formed case is pinned as hard as
+  the broken ones.
+- **Verified against agy 1.1.11, no change needed:** all three tiers still resolve
+  (`flash` → `gemini-3.5-flash-high`, `flash-lo` → `-low`, `pro` → `gemini-3.1-pro-high`);
+  the exit-14 classifier still fires on an unknown model; and 1.1.11's reworded
+  model-loading errors do not disturb the exit-15 path, whose message still contains
+  `auto-denied` and `permissions.allow` verbatim — two independent anchors the classifier
+  already matches.
+  **`/usage` and `/quota` were considered for `doctor` and rejected on evidence:** they
+  return zero bytes here, which is not a bug but this account type — 1.1.11's own notes say
+  credits do not apply to accounts signed in through a Google Cloud project or ADC.
+  `/model`, `/effort` and `/skills` all return data on the same setup, which is what made
+  the probe above possible.
+- **The suite had a false-negative construct in 25 assertions, and this release woke it
+  up.** `printf '%s' "$x" | grep -q PAT` is the shape fixed in 0.21.1 and explained in a
+  comment in this very file: `grep -q` exits at the first match and closes the pipe, the
+  writer dies of SIGPIPE (141), and `set -o pipefail` marks the whole pipeline failed — so
+  the assertion reads "not found" while the text is right there. The window is whatever the
+  writer still has to emit *after* the matched line, which is why it survives review and
+  why adding output below a match can revive it: the 1.1.9 version-gate assertion started
+  failing 1–4 times in 8 concurrent runs on this branch and never on master.
+  Diagnosed rather than guessed. Instrumenting the *test* made it vanish (it changes the
+  timing), so the trace went inside `doctor` instead: `mktemp` never failed and `AGY_VER`
+  was never empty, which left only the assertion misreading output that had in fact been
+  printed. All 25 now use a `case`-based `has()` — same test, no second process, no pipe —
+  and the five `sed … | grep -q` source scans give `grep` a process substitution instead,
+  so `sed`'s death is no longer `pipefail`'s business. Those five failed **silently**: a
+  false negative there reads as "the property holds", and among them are the guards for
+  the delegate wrapper's `--help` probe, `agy_guard`'s pipe rule, and the `sort -V`
+  dependency. 24 consecutive clean runs at the concurrency that reproduced it; each
+  converted scan re-checked by mutation.
+  **That fix then shipped the same defect in a new shape, and both reviewers caught it.**
+  `has()` was defined beside the doctor tests, above which two call sites already sat —
+  and bash does not hoist, so those two were `command not found`, exit 127, `else` taken
+  unconditionally, `ok` printed regardless. One of them was the "json envelope leaked to
+  stdout" assertion, which had already been voided once before by a different accident.
+  Helpers now live at the top of the file, and `tests/check-helper-order.py` runs first
+  and fails the suite if any function is called above its definition. bash 4's
+  `command_not_found_handle` was tried for this and **removed**: macOS ships bash 3.2,
+  where merely defining it is a silent no-op — a guard that reads as protection and
+  provides none, which is the defect this whole entry is about. The static check works on
+  any shell and was verified by putting the original bug back: it names the call site and
+  the definition line. Its first regex missed `elif`, a call inside a `case` branch, and a
+  brace group — all three confirmed against the old pattern — so it now splits the line
+  into command segments and compares each segment's first word, with no list of contexts
+  to keep complete. The checker has its own tests, including a no-false-positive case,
+  because a guard that misses a shape is the defect it exists to prevent. It also flagged
+  its own test data, which is fair: a fixture written across real lines is not
+  distinguishable from code, so the fixtures are single-line now.
+- **The exit-15 message now points at the rule itself.** It is the one place someone
+  actually lands when a write is soft-denied, and it recommended `permissions.allow`
+  without allowing that the rule *is* the problem: "if a rule is ALREADY in that file and
+  you are still reading this, suspect the rule: run `agy-doctor`". Same caveat added to
+  `agents/antigravity-delegate.md` and `commands/delegate.md`, the two operational files
+  consulted while building a delegation call. The pre-run write nudge got the short half only — `<dir>` is a placeholder, `agy-doctor` will say whether yours parses — because that one fires on every write-looking task, while the diagnosis belongs where the failure is.
+  **And the same overclaim the code fix removed was still sitting in the prose it was
+  copied from.** Five documents put the placeholder caveat next to the pre-1.1.11
+  match-everything history, which reads as though a mistyped `write_file(<dir>)` carried
+  it — the exact conflation `bad_allow_rules` classifies apart (`unparseable`, not
+  `zerowords`). Fixing the output and leaving the sentence that produced it just relocates
+  the error, so all five now say the placeholder grants nothing on any version and name
+  `command(...)` explicitly for the part that is version-sensitive.
+- **The remaining 24 piped assertions are converted too, because the direction that was
+  left is the dangerous one.** Review flagged them as a follow-up on the grounds that they
+  are positive matches, where a SIGPIPE produces a noisy false FAIL. A scan says the
+  opposite: 17 of them are *negative* assertions, and SIGPIPE is only possible when
+  `grep -q` finds a match and exits early — which for a negative assertion is exactly the
+  moment it is supposed to fail. They cannot flake in CI, because they only break when
+  they have a real bug to report. Each becomes `grep … <<<"$x"`: a here-string has no
+  writer process to kill and the pipeline is one command, so `pipefail` has nothing to
+  poison, and `-F` / `-i` / `-E` / BRE semantics survive untouched — unlike a rewrite to
+  `case`. Verified by mutation on both the plain and the `-qE` alternation forms.
+  The first pass at this shipped a regex that truncated a grep pattern at a `)` inside it,
+  producing valid shell that searched for the wrong string and hung a polling loop. Caught
+  because the suite stopped completing, not because anything reported it.
+
 ## 0.22.4
 - **`--tier` did nothing on agy below 1.1.10, and nothing said so.** agy 1.1.10 fixed
   `--model` and `--effort` being *ignored in headless `-p`* — the flag was applied after

README.md

@@ -202,6 +202,13 @@ Delegation doesn't save money by itself — these do (also in the skill):
   - **`permissions.allow` in `~/.gemini/antigravity-cli/settings.json`** — a
     `write_file(<dir>)` entry allows writes **recursively beneath `<dir>`** and needs no
     flag. This is the narrower grant and usually the right one.
+    **`<dir>` is a placeholder — substitute a real path.** Left as written it grants
+    nothing on any agy version, and the write is soft-denied with the rule sitting visibly
+    in the file. A *different* mistake is the version-sensitive one: a `command(...)` rule
+    that names no command (`command(time)`, a comment-only entry, `()`) matched **every**
+    command before agy 1.1.11 and silently auto-approved anything the agent ran — broader
+    than the `--yolo` it was chosen instead of. 1.1.11 makes that entry match nothing too.
+    `agy-doctor` checks your entries and reports the consequence that actually applies.
   - **`--yolo`** (`--dangerously-skip-permissions`) — auto-approves **all** tools, not just
     writes. Needed when no rule covers the target, and for web / Vertex AI Search / terminal
     tools.

agents/antigravity-delegate.md

@@ -87,7 +87,10 @@ Options: `--tier flash|flash-lo|pro` · `--dir <repo-root>` (so agy reads
   write needs a grant. Pass `--yolo` unless the user has a `permissions.allow`
   `write_file(<dir>)` rule covering the target in `~/.gemini/antigravity-cli/settings.json`
   — that grants the write recursively beneath `<dir>` with no flag, and is narrower than
-  `--yolo`, which approves every tool. You cannot see that file, so `--yolo` stays the
+  `--yolo`, which approves every tool. If they say a rule is in place and the write is
+  still soft-denied, have them run `agy-doctor` before anything else: an entry agy cannot
+  parse grants nothing. (The "granted everything before 1.1.11" history belongs to a
+  `command(...)` rule naming no command, not to a mistyped `write_file()`.) You cannot see that file, so `--yolo` stays the
   default; if a run comes back exit `15`, the allow-rule is the smaller fix. Either way tell
   the caller to run on a dedicated branch/worktree and review the diff before merging.
 - **Read-only** (analysis, first-pass review, search): no `--yolo` needed unless

commands/delegate.md

@@ -14,7 +14,8 @@ Do this:
    **If the task WRITES files or uses tools** (web / Vertex AI Search / terminal), it needs
    a grant. For a plain file write the narrower one is a `write_file(<dir>)` entry under
    `permissions.allow` in `~/.gemini/antigravity-cli/settings.json` (recursive beneath
-   `<dir>`, no flag needed). Otherwise pass **`--yolo`**, which auto-approves all tools and
+   `<dir>`, no flag needed — substitute a real path for `<dir>`; if a rule is already
+   there and the write is still denied, `agy-doctor` checks whether agy can parse it). Otherwise pass **`--yolo`**, which auto-approves all tools and
    is what web / Vertex AI Search / terminal need. Without a grant,
    headless agy leaves your workspace untouched while still reporting success (it
    describes / scratch-diverts / soft-denies depending on version; issue #10). `--mode

docs/POC-PLAYBOOK.md

@@ -81,6 +81,13 @@ After each lever: rerun the task → rerun the gate → keep only if quality hel
   - **`permissions.allow` in `~/.gemini/antigravity-cli/settings.json`** — a
     `write_file(<dir>)` entry allows writes **recursively beneath `<dir>`** and needs no
     flag. This is the narrower grant and usually the right one.
+    **`<dir>` is a placeholder — substitute a real path**, and run `agy-doctor` before a
+    measured run. Left as written the rule grants nothing on any version, so the write is
+    soft-denied with the rule sitting visibly in the file. A separate mistake — a
+    `command(...)` rule naming no command — matched **every** command before **agy
+    1.1.11**, silently auto-approving anything the agent ran. Either one makes an arm's
+    write behaviour a second uncontrolled variable, which is exactly what §5 says a cost
+    claim cannot carry.
   - **`--yolo`** (`--dangerously-skip-permissions`) — auto-approves **all** tools, not just
     writes. Needed when no rule covers the target, and for web / Vertex AI Search / terminal
     tools.

docs/TROUBLESHOOTING.md

@@ -110,6 +110,14 @@ run still "succeeds"** ([#10](https://github.com/yuting0624/antigravity-for-clau
   This is the rule agy's own soft-deny message is naming. Confirmed on agy 1.1.9 by a
   controlled A/B ([#37](https://github.com/yuting0624/antigravity-for-claude-code/issues/37));
   a glob form (`write_file(/path/**)`) was reported *not* to match.
+  **Substitute a real path for `<dir>`** — and if the rule is in place and the write is
+  *still* soft-denied, suspect the rule before suspecting agy. An entry agy cannot parse
+  grants nothing on any version, which is exactly this exit 15 with the rule sitting
+  right there in the file. Only one shape of mistake is version-sensitive, and it is not
+  this one: a `command(...)` rule naming no command — `command(time)`, a comment-only
+  entry, `()` — matched **every** command before **1.1.11** and silently auto-approved
+  anything the agent ran. Run `agy-doctor`: it validates each entry and reports the
+  consequence that actually applies to yours.
 - **Or pass `--yolo`** (`--dangerously-skip-permissions`) — works across all agy versions,
   but auto-approves **all** tools, not just the write. Required anyway for web / Vertex AI
   Search / terminal when no rule covers them. (`--mode accept-edits` only wrote headless on

scripts/agy-delegate.sh

@@ -253,7 +253,7 @@ if [ "$YOLO" -eq 0 ] && [ "$PRINT_CMD" -ne 1 ]; then
   shopt -s nocasematch
   case "$PROMPT" in
     *implement*|*scaffold*|*migrate*|*refactor*|*"write the file"*|*"create the file"*|*"edit the file"*)
-      echo "agy-delegate: note: this looks like a write task and --yolo is not set. Headless agy will NOT touch your workspace without a write grant (it describes / scratch-diverts / soft-denies depending on version, while the run still 'succeeds'; issue #10). Two grants work: a permissions.allow rule matching the target — write_file(<dir>), a recursive prefix, in ~/.gemini/antigravity-cli/settings.json — which is the narrower one and needs no flag; or --yolo, which auto-approves ALL tools. If a rule already covers your target, ignore this. Otherwise add one, or pass --yolo on a dedicated branch, and verify with git status." >&2 ;;
+      echo "agy-delegate: note: this looks like a write task and --yolo is not set. Headless agy will NOT touch your workspace without a write grant (it describes / scratch-diverts / soft-denies depending on version, while the run still 'succeeds'; issue #10). Two grants work: a permissions.allow rule matching the target — write_file(<dir>), a recursive prefix, in ~/.gemini/antigravity-cli/settings.json — which is the narrower one and needs no flag; or --yolo, which auto-approves ALL tools. If a rule already covers your target, ignore this — but <dir> is a placeholder, and agy-doctor will tell you whether yours actually parses. Otherwise add one, or pass --yolo on a dedicated branch, and verify with git status." >&2 ;;
   esac
   shopt -u nocasematch
 fi
@@ -494,7 +494,7 @@ if [ -z "${OUT//[$' \t\n\r']/}" ]; then
     *"auto-denied"*|*"permissions.allow"*|*"permission that headless"*|*"dangerously-skip-permissions"*)
       shopt -u nocasematch
       [ -s "$ERR" ] && cat "$ERR" >&2
-      echo "agy-delegate: agy soft-denied a tool that needs permission (headless can't prompt) — no work was done. For a FILE WRITE, the narrower fix is a permissions.allow rule covering the target in ~/.gemini/antigravity-cli/settings.json — write_file(<dir>) matches recursively beneath <dir> — which needs no flag; --yolo also works but auto-approves ALL tools. Other tools (web / Vertex AI Search / terminal) need --yolo unless a rule covers them. agy's own message above names the specific permission it wanted. (agy >= 1.1.3)" >&2
+      echo "agy-delegate: agy soft-denied a tool that needs permission (headless can't prompt) — no work was done. For a FILE WRITE, the narrower fix is a permissions.allow rule covering the target in ~/.gemini/antigravity-cli/settings.json — write_file(<dir>) matches recursively beneath <dir> — which needs no flag; --yolo also works but auto-approves ALL tools. Other tools (web / Vertex AI Search / terminal) need --yolo unless a rule covers them. agy's own message above names the specific permission it wanted. If a rule is ALREADY in that file and you are still reading this, suspect the rule: run agy-doctor, because an entry agy cannot parse grants nothing (and before agy 1.1.11 granted everything). (agy >= 1.1.3)" >&2
       signal PERMISSION_DENIED "agy soft-denied a permissioned tool in headless — add a permissions.allow rule or pass --yolo"; exit 15 ;;
   esac
   shopt -u nocasematch