fix(subagents): break recursion — disarm user hooks for subagent sessions#53
Merged
Conversation
…ions
Each ``claude -p`` capture-extract / dream / graph_extraction subagent
is itself a Claude Code session. The user's PostToolUse / Stop /
SessionEnd hooks therefore fire for the subagent's *own* tool uses.
Without a guard, this cascades:
1. Active session does N tool uses → PostToolUse hook → spawns
capture-extract subagent.
2. Subagent's own tool uses (Read, Bash, Grep …) re-fire PostToolUse,
which writes to *its* session_id buffer.jsonl.
3. After ``LORE_CAPTURE_N`` entries, the hook spawns *another*
capture-extract for the subagent — and so on.
Observed in production for one user: ~700 spawns/hour, ~$34/h on
Haiku, 685 distinct sessions hit in 60 minutes despite only one
interactive ``claude`` process running. ``--strict-mcp-config`` from
PR #52 isolates MCP but does not isolate hooks.
Two layers of defense:
1. ``subagent_config.settings_body()`` now writes ``hooks`` as
empty arrays. ``--settings`` overrides the user's hooks for
the subagent's session.
2. New ``SubagentConfig.env_overrides()`` returns
``LORE_AUTO_SAVE=false`` and ``LORE_DREAM_AUTO=false``. The
hook scripts honor these as master kill switches and exit 0
immediately. All three spawn sites now pass
``env={**os.environ, **cfg.env_overrides()}`` so the guard
survives any caching of ``settings.json`` by the running
parent Claude Code process.
3 new tests: hooks-empty in materialized settings, env_overrides
shape, env_overrides parity across roles. Existing
TestSpawnClaudeArgs extended to assert recursion-guard env vars
are passed to subprocess.Popen.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Follow-up to #52. The cheap-subagent flag set isolates MCP and pins the model to Haiku/Sonnet, but does not isolate hooks — and that turns out to be the dominant cost path.
Each
claude -psubagent is itself a Claude Code session, so the user'sPostToolUse/Stop/SessionEndhooks fire for the subagent's own tool uses. The lore-capture-* hooks then spawn nested capture-extracts for the subagent's session_id, which spawn more, etc.Observed in production on a single user's machine:
claudeprocess runningThe 685-distinct-sessions number is what gives it away — none of those sessions exist in the user's interactive workflow; they're subagent spawns each opening their own Claude Code session, accumulating their own buffer.jsonl, and triggering their own hooks.
Fix — two layers of defense
subagent_config.settings_body()now writeshooksas empty arrays for all four event names.--settingsoverrides the user's hooks for the subagent.SubagentConfig.env_overrides()returnsLORE_AUTO_SAVE=falseandLORE_DREAM_AUTO=false. The hook scripts honor these as master kill switches and exit 0 immediately. All three spawn sites passenv={**os.environ, **cfg.env_overrides()}so the guard survives even if the parent Claude Code process has cached~/.claude/settings.json.Test plan
tests/test_subagent_config.py::TestSettings::test_hooks_are_empty_to_break_subagent_recursiontests/test_subagent_config.py::TestEnvOverrides— shape + role paritytests/services/test_graph_extraction.py::TestSpawnClaudeArgs— extended to assert env vars onsubprocess.Popenpytest tests/ --ignore=tests/integration→ 2749 passed, 0 failedruff checkclean🤖 Generated with Claude Code