Repository navigation
v0.7.0 install/update wires the legacy blocking PreToolUse gate, not hook_augment (regresses #214) #409
Description
Summary
On a clean v0.7.0 install -y --standard (and update), the Claude Code PreToolUse hook written to ~/.claude/hooks/cbm-code-discovery-gate is the old blocking shell gate (exit 2, gate file keyed on bare $PPID) — not the non-blocking augmenter shipped in the binary (src/cli/hook_augment.c, commit c29e6d5). So the exact symptom of #214 returns on every install/update, even though #214 is marked fixed in v0.7.0.
Why it misbehaves (same root as #214)
$PPID is the shell spawned for each hook invocation, so it rotates per tool call. The gate file /tmp/cbm-code-discovery-gate-$PPID therefore never matches a prior one → the "first call per session" check is always true → every Grep/Glob/Read is blocked with exit 2, not just the first.
The installed hook (v0.7.0):
#!/bin/bash
# Gate hook: nudges Claude toward codebase-memory-mcp for code discovery.
# First Grep/Glob/Read/Search per session -> block. Subsequent -> allow.
# PPID = Claude Code process PID, unique per session.
GATE=/tmp/cbm-code-discovery-gate-$PPID
...
touch "$GATE"
echo 'BLOCKED: ...' >&2
exit 2Repro
codebase-memory-mcp install -y --standard(orupdate) on v0.7.0, Linux, Claude Code.- In a session, issue several
Read/Grepcalls. - Each is blocked by the gate (not nudged once) — the
claudeprocess reads the hook fresh each call and$PPIDdiffers every time.
Expected
install/update should wire the non-blocking augmenter (the hook_augment path that's already in the binary), so Read is no longer gated and the nudge is emitted at most once per session — the behaviour #214 was closed as delivering.
Note
The binary genuinely has the fix; this is a packaging/install gap — the install template still emits the legacy blocking script. A session_id-keyed (from the PreToolUse stdin JSON) non-blocking augmenter that exit 0s avoids both the $PPID rotation and any ps/process-name assumptions.
Environment
- v0.7.0, Linux (cgroup v2), Claude Code, standard stdio variant.
Activity
DeusData commentedon Jun 6, 2026
Thanks for the detailed write-up — the
$PPIDanalysis is spot-on for the old gate.I'm having trouble reproducing this on a clean
v0.7.0. Onmain(and inv0.7.0, which already contains the augmenter fixc29e6d5), bothinstallandupdatewrite~/.claude/hooks/cbm-code-discovery-gateas the non-blocking augmenter shim, not the old gate:No
$PPID, noexit 2;settings.json'sPreToolUsecommand points at that script; the file is written withfopen(..., "w")(truncate, so it overwrites any older gate); and a regression test guards it (cli_hook_gate_script_no_predictable_tmp_issue384— asserts noPPID, delegates tohook-augment).To pin down what's different in your environment, could you share:
~/.claude/hooks/cbm-code-discovery-gate.codebase-memory-mcp --version.PreToolUseblock (matcher +command) from~/.claude/settings.json.v0.7.0version (roughly which)?Leading suspicion: a stale
PreToolUseentry insettings.jsonfrom an older install still pointing at an inline/$PPIDgate rather than the rewritten shim — (3) + (4) would confirm it.test(repro): install/UX reproductions for #570, #409, #431
DeusData commentedon Jun 26, 2026
This is fixed. The installer now wires the non-blocking
hook-augmentshim rather than the legacy blockingPreToolUsegate, and the gate-script writer truncates and rewrites the non-blocking content on upgrade (so a stale blocking script from a pre-0.7.0 install is replaced). We've added a regression guard for the upgrade/overwrite path specifically.Closing as fixed — reopen if a fresh install or upgrade still wires a blocking gate for you. Thanks!