Hooks
What fable-mode runs automatically, and when. A hook is a command Claude Code fires at a fixed moment, without you asking for it.
> /plugin marketplace add rennf93/opus-fable-playbook > /plugin install fable-mode@opus-fable-playbook
Ships with fable-mode. Installing the plugin gets these hooks.
What fires, and when
SessionStart
Fires once when a session begins, and again after a context compaction. It is where a plugin sets up its environment, or restores state the compaction dropped.
- Matches
startup|clear|compact|resume"${CLAUDE_PLUGIN_ROOT}/hooks/session-start.sh"
UserPromptSubmit
Fires before Claude sees each prompt you send. A plugin can use it to inject context, so the same instruction reaches the model every turn instead of only at session start.
"${CLAUDE_PLUGIN_ROOT}/hooks/prompt-nudge.sh"
Stop
"${CLAUDE_PLUGIN_ROOT}/hooks/stop-gate.sh"
SubagentStop
"${CLAUDE_PLUGIN_ROOT}/hooks/stop-gate.sh" subagent
PreToolUse
- Matches
Bash"${CLAUDE_PLUGIN_ROOT}/hooks/bash-discipline.sh"
PostToolUse
- Matches
Bash"${CLAUDE_PLUGIN_ROOT}/hooks/honesty-nudge.sh"
PreCompact
- Matches
manual|auto"${CLAUDE_PLUGIN_ROOT}/hooks/precompact.sh"
Where it lives
- hooks/bash-discipline.shRunsGitHub
Read the script
#!/usr/bin/env bash # PreToolUse[Bash]: deny pure shell file-reads; dedicated tools exist. set -u DIR="$(cd "$(dirname "$0")" && pwd)" # shellcheck source=hooks/lib/telemetry.sh . "$DIR/lib/telemetry.sh" INPUT="$(cat)" || exit 0 CMD="$(printf '%s' "$INPUT" | python3 -c \ 'import json,sys; print(json.load(sys.stdin).get("tool_input",{}).get("command",""))' \ 2>/dev/null || true)" [ -z "$CMD" ] && exit 0 # Pipelines, compounds, redirects, heredocs are legitimate — allow. printf '%s' "$CMD" | grep -qE '\||&&|;|>|<<' && exit 0 DENY=0 printf '%s' "$CMD" | grep -qE '^[[:space:]]*(cat|head|tail|less|more)[[:space:]]' && DENY=1 printf '%s' "$CMD" | grep -qE '^[[:space:]]*sed[[:space:]]+-n[[:space:]]' && DENY=1 [ "$DENY" -eq 0 ] && exit 0 SESSION="$(printf '%s' "$INPUT" | python3 -c \ 'import json,sys; print(json.load(sys.stdin).get("session_id","unknown"))' \ 2>/dev/null || true)" fable_telemetry "bash-discipline" "shell-read" "$SESSION" cat <<'JSON' {"hookSpecificOutput": {"hookEventName": "PreToolUse", "permissionDecision": "deny", "permissionDecisionReason": "Fable tool discipline: use the dedicated Read/Grep tools instead of shell file-reads (cat/head/tail/less/sed -n). Read is paginated and line-numbered; Grep searches without loading whole files."}} JSON exit 0 - hooks/honesty-nudge.shRunsGitHub
Read the script
#!/usr/bin/env bash # PostToolUse[Bash]: when output shows failures, nudge verbatim reporting. set -u DIR="$(cd "$(dirname "$0")" && pwd)" # shellcheck source=hooks/lib/telemetry.sh . "$DIR/lib/telemetry.sh" INPUT="$(cat)" || exit 0 RESP="$(printf '%s' "$INPUT" | python3 -c \ 'import json,sys; print(json.dumps(json.load(sys.stdin).get("tool_response","")))' \ 2>/dev/null || true)" [ -z "$RESP" ] || [ "$RESP" = '""' ] && exit 0 HIT=0 printf '%s' "$RESP" | grep -qE 'FAILED |= FAILURES =|test result: FAILED|--- FAIL|AssertionError|Traceback \(most recent call last\)' && HIT=1 printf '%s' "$RESP" | grep -qE 'Tests:[^"]*failed' && HIT=1 [ "$HIT" -eq 0 ] && exit 0 SESSION="$(printf '%s' "$INPUT" | python3 -c \ 'import json,sys; print(json.load(sys.stdin).get("session_id","unknown"))' \ 2>/dev/null || true)" fable_telemetry "honesty-nudge" "failure-output" "$SESSION" cat <<'JSON' {"hookSpecificOutput": {"hookEventName": "PostToolUse", "additionalContext": "A command just reported failures. Fable honesty rule: report this outcome verbatim (the actual failing output) in your final message; do not summarize it as mostly-working or claim success."}} JSON exit 0 - hooks/precompact.shRunsGitHub
Read the script
#!/usr/bin/env bash # PreCompact: shape what survives compaction. set -u cat > /dev/null || true cat <<'EOF' Compaction guidance (fable-mode): the summary must preserve, outcome-first: (1) current task state and remaining work, (2) what was verified, with the actual results, (3) any failures not yet reported to the user, verbatim, (4) pending user decisions, (5) paths of files being modified. EOF exit 0
- hooks/prompt-nudge.shRunsGitHub
Read the script
#!/usr/bin/env bash # UserPromptSubmit: <=40-token doctrine nudge; question-shape heuristic. set -u INPUT="$(cat)" || exit 0 PROMPT="$(printf '%s' "$INPUT" | python3 -c \ 'import json,sys; print(json.load(sys.stdin).get("prompt",""))' 2>/dev/null || true)" [ -z "$PROMPT" ] && exit 0 case "$PROMPT" in /*) exit 0 ;; esac TRIMMED="$(printf '%s' "$PROMPT" | sed 's/[[:space:]]*$//')" FIRST="$(printf '%s' "$PROMPT" | awk '{print tolower($1); exit}')" LOWER="$(printf '%s' "$TRIMMED" | tr '[:upper:]' '[:lower:]')" case "$TRIMMED" in *\?) Q=1 ;; *) Q=0 ;; esac case "$FIRST" in why|what|how|is|does|should|can|are|do|where|when|who|which) Q=1 ;; esac case "$LOWER" in # imperative-investigate-then-report prompts ("run the tests and tell # me where this project stands") are assess-only even though they # don't start with a question word or end in "?". *where*stand*) Q=1 ;; esac if [ "${Q:-0}" = "1" ]; then printf 'This prompt is question-shaped: deliver your assessment; do not change code unless asked.' else printf 'Fable reminders: lead the final message with the outcome; finish work instead of narrating it; parallelize independent tool calls; delegate broad searches.' fi printf '\n' exit 0 - hooks/session-start.shRunsGitHub
Read the script
#!/usr/bin/env bash # SessionStart: inject the doctrine card; flag inactive output style. set -u DIR="$(cd "$(dirname "$0")" && pwd)" cat > /dev/null || true # drain stdin CARD="$DIR/lib/doctrine-card.md" [ -f "$CARD" ] && cat "$CARD" STYLE="$(python3 -c "import json,os;print(json.load(open(os.path.expanduser('~/.claude/settings.json'))).get('outputStyle',''))" 2>/dev/null || true)" if [ "$STYLE" != "Fable" ]; then printf '\nNote: the Fable output style is not set in user settings. If this session should run fable-mode fully, suggest the user run /output-style fable (or set "outputStyle": "Fable" in settings).\n' fi exit 0 - hooks/stop-gate.shRunsGitHub
Read the script
#!/usr/bin/env bash # Stop/SubagentStop gate: block turn endings that promise instead of do. # Usage: stop-gate.sh [subagent] Fail-open: any internal error => exit 0. set -u DIR="$(cd "$(dirname "$0")" && pwd)" # shellcheck source=hooks/lib/telemetry.sh . "$DIR/lib/telemetry.sh" INPUT="$(cat)" || exit 0 py() { printf '%s' "$INPUT" | python3 -c "$1" 2>/dev/null || true; } ACTIVE="$(py 'import json,sys; print(json.load(sys.stdin).get("stop_hook_active", False))')" [ "$ACTIVE" = "True" ] && exit 0 SESSION="$(py 'import json,sys; print(json.load(sys.stdin).get("session_id","unknown"))')" LAST="$(printf '%s' "$INPUT" | python3 "$DIR/lib/last_message.py" 2>/dev/null)" || exit 0 [ -z "$LAST" ] && exit 0 # Final paragraph = last blank-line-separated block (awk paragraph mode). FINAL="$(printf '%s' "$LAST" | awk -v RS='' 'END{print}')" [ -z "$FINAL" ] && exit 0 VERBS='(start|begin|proceed|continue|create|implement|write|update|fix|add|run|check|investigate|work|make|set|move|look|open|draft|explore|apply|push|refactor|clean|test)' MATCH="" printf '%s' "$FINAL" | grep -qiE "(^|[^a-z])i('|’)?ll (now |then |next |also |go ahead and )?$VERBS" && MATCH="ill-promise" [ -z "$MATCH" ] && printf '%s' "$FINAL" | grep -qiE "(^|[^a-z])i will (now |then |next |also )?$VERBS" && MATCH="i-will" [ -z "$MATCH" ] && printf '%s' "$FINAL" | grep -qiE "(^|[[:space:]])next steps?:" && MATCH="next-steps" [ -z "$MATCH" ] && printf '%s' "$FINAL" | grep -qiE "let me know (if|when|whether|what|which|and)" && MATCH="let-me-know" [ -z "$MATCH" ] && printf '%s' "$FINAL" | grep -qiE "would you like me to" && MATCH="would-you-like" [ -z "$MATCH" ] && printf '%s' "$FINAL" | grep -qiE "(^|[^a-z])shall i " && MATCH="shall-i" # Golden calibration 2026-07-02: bare "want me to " blocked real Fable endings # (assess-only tasks ending "want me to apply the fix?" — a genuine decision # question). Anchor to continuation verbs so only in-scope deferral blocks. [ -z "$MATCH" ] && printf '%s' "$FINAL" | grep -qiE "(^|[^a-z])want me to (continue|proceed|keep going|finish|do the rest)" && MATCH="want-me-to" MODE="${1:-main}" # Opt-in LLM judge tier (main mode only, only when tier 1 found nothing). if [ -z "$MATCH" ] && [ "$MODE" = "main" ] && [ "${FABLE_STOP_JUDGE:-0}" = "1" ] \ && command -v claude >/dev/null 2>&1; then # --bare requires API-key auth and breaks OAuth-only machines (Task 15 finding); inherit session auth instead — plugin contamination is acceptable for a 10-word verdict. VERDICT="$(printf 'Does this assistant turn-ending violate the rule "finish the work instead of promising it; do not seek permission for reversible in-scope actions"? Reply with exactly YES or NO.\n\n---\n%s' "$FINAL" \ | claude -p --model "${FABLE_STOP_JUDGE_MODEL:-claude-haiku-4-5-20251001}" 2>/dev/null | tr -d '[:space:]')" [ "$VERDICT" = "YES" ] && MATCH="judge" fi [ -z "$MATCH" ] && exit 0 if [ "$MODE" = "subagent" ]; then fable_telemetry "stop-gate-subagent" "$MATCH" "$SESSION" REASON="Fable subagent discipline: your final message is your return value. Return your findings now — conclusions with evidence, not intentions, plans, or offers." else fable_telemetry "stop-gate" "$MATCH" "$SESSION" REASON="Fable turn discipline: your last paragraph promises or proposes work instead of doing it. Do that work now — retry errors and gather missing information yourself. If you are genuinely blocked on something only the user can provide, state that blocking question plainly and stop." fi printf '{"decision": "block", "reason": "%s"}' "$REASON" exit 0
Read the script before you install anything that runs on your machine. This is the one part of a plugin that acts without being asked.
Make Claude Opus 4.8 in Claude Code behave as much like Claude Fable 5 as possible. The doctrine was transcribed by Fable 5 itself; hooks enforce it at the harness level; an eval loop measures convergence against golden Fable transcripts.

