distill

<objective>

Analyze how Claude Code is used, surface concrete improvements — new agents/skills to reduce repetition, or consolidate lessons into governance files (rules, agent instructions, skill updates) — without duplicating what exists.

NOT for single-file edits or quality checks — use /foundry:audit for config quality checks. NOT for audit-only scan for extraction candidates (use /foundry:audit --efficiency instead of distill executables for detection-only).

</objective> <inputs>

• $ARGUMENTS: optional. Modes:
  • Omitted: analyze existing patterns and agents; generate suggestions proactively.
  • prune [--eager] — evaluate project memory file for stale, redundant, or verbose entries. Default: advisory diff + apply prompt. --eager: score every entry (Usage likelihood × Impact → Tier P0/P1/P2), print full scored table with # column, let user select by tier or item numbers, delegate edits to foundry:curator.
  • memory [--eager] — read .notes/lessons.md and memory feedback files, distill recurring patterns into proposed rule files, agent instruction updates, skill workflow changes. --eager: include Pattern count, Strength, Tier columns in proposal table; let user select clusters to promote by tier or item numbers; delegate writes to foundry:curator.
  • external <source> [--eager] — analyse external plugin, skill, or agentic resource, produce structured adoption proposal. <source> is URL, file path, or local directory. --eager: lower adoption bar — recommend partial adoption even for single useful components.
  • executables [--eager] [<run-dir-or-report-path>] — perform bin/ extraction from /foundry:audit --efficiency Check 33 candidates. Auto-detects latest run dir under .reports/audit/; pass optional path to target specific run dir or report file. Runs inline Check 33 scan when no report exists. Default gates on HIGH/MEDIUM verdict. --eager: also surface LOW verdict clusters as extraction candidates. Spawns foundry:sw-engineer per cluster. Skip to Mode: Executables Extraction below.
  • [--eager] <recurring task description> — use description as context when generating suggestions. --eager: lower frequency threshold from 3+ to 2+ occurrences; single high-effort occurrence also qualifies.
  • --project — in prune and memory modes, show interactive project picker: enumerate all slugs under ~/.claude/projects/*/memory/ with MEMORY.md size in tokens, let user select which project(s) to operate on. Omit to operate across all projects automatically. No effect on other modes.

</inputs> <compaction> - Key boundary 1: end of Step 2 frequency heuristics (default mode only — prune/memory/external/executables exit early), before Step 3 gap analysis. - Preserve at boundary 1: EAGER flag, ARGUMENTS (stripped), no run-dir (default mode is stateless). - Terminal paths: end of Step 5 report (default mode); end of Memory Pruning mode (both eager and standard branches). </compaction> <workflow>

Task hygiene — task tools may be deferred; load before first use: ToolSearch(query="select:TaskList,TaskCreate,TaskUpdate,TaskGet", max_results=4). Call TaskList first and triage each task it returns: completed if work clearly done, deleted if orphaned, keep in_progress only if genuinely continuing. Never spend a turn on bookkeeping alone — every TaskCreate/TaskUpdate ships in the same response as the next substantive tool call; one exception, TaskUpdate(completed) immediately before a long output block (rules/task-lifecycle.md).

export CSID="${CLAUDE_CODE_SESSION_ID:-$PPID}"
python "${CLAUDE_PLUGIN_ROOT:-plugins/cc_foundry}/bin/extract-keep-flag.py" distill-state "$ARGUMENTS" --out-file "${TMPDIR:-/tmp}/distill-state-${CSID}/keep-items"  # timeout: 5000 — parses --keep, clears stale contract, makes state dir
eval "$(python "${CLAUDE_PLUGIN_ROOT:-plugins/cc_foundry}/bin/parse-skill-flags.py" --flags eager "$ARGUMENTS")"  # timeout: 5000
EAGER="$FLAG_EAGER"
ARGUMENTS="$CLEAN_ARGS"
echo "EAGER=$EAGER"  # shell vars don't persist across Bash calls — read from stdout
echo "ARGUMENTS_STRIPPED=$ARGUMENTS"

Note: EAGER and stripped ARGUMENTS are set by this Bash block, but shell variable state does not persist across separate Bash() tool calls. After this block runs, read its stdout (EAGER=true/false, ARGUMENTS_STRIPPED=...), carry those values as model-context references for all subsequent mode dispatch and threshold decisions. Do not rely on $EAGER as a live shell variable in later steps — substitute the literal boolean value read from stdout.

eval "$(python "${CLAUDE_PLUGIN_ROOT:-plugins/cc_foundry}/bin/parse-skill-flags.py" --flags project "$ARGUMENTS")"  # timeout: 5000
PROJECT_FLAG="$FLAG_PROJECT"
ARGUMENTS="$CLEAN_ARGS"
echo "PROJECT_FLAG=$PROJECT_FLAG"
echo "ARGUMENTS_FINAL=$ARGUMENTS"

Note: PROJECT_FLAG does not persist across Bash calls. Read its value from stdout line PROJECT_FLAG=true/false, carry as model-context reference. When true, mode must run interactive picker before operating.

<!-- ARCH.md beside this file diagrams the runs, gates and parallel fan-out. Documentation only, never loaded — update it in the same commit as any change to step order, gate placement, or agent fan-out. -->

Step 1: Inventory existing agents and skills

Use Glob tool to enumerate agents and skills across all sources — project-local AND plugin-namespaced — avoids false-gap findings when candidate already exists in plugin:

• Project-local: pattern agents/*.md, path .claude/; pattern skills/*/SKILL.md, path .claude/
• Plugin source (workspace): pattern */agents/*.md, path plugins/; pattern */skills/*/SKILL.md, path plugins/
• Installed plugin cache (if accessible): resolve cache root — PLUGIN_CACHE="${CLAUDE_PLUGIN_ROOT:-plugins/cc_foundry}" — then use Glob tool on $PLUGIN_CACHE for pattern */agents/*.md and */skills/*/SKILL.md

For each agent/skill found, extract: name, description, tools, purpose. Tag each entry with plugin namespace (e.g. foundry:sw-engineer, oss:resolve) — used in Step 3 gap analysis to prevent recommending duplicates of plugin-namespaced agents/skills.

Step 2: Analyze work patterns

Mode-token normalization — all mode dispatches below compare against the first whitespace-delimited token of stripped ARGUMENTS (after --eager removal). Use this single rule consistently; don't rely on exact equality of the full $ARGUMENTS string — trailing flags/spaces from prior parsing may differ.

--project scope check (run before the mode checks below): if PROJECT_FLAG (read from stdout above) is true and the first token is not prune or memory, print ⚠ `--project` ignored — it applies only to `prune` and `memory` modes. once and repeat it in the final reply (a mid-run print can arrive as an empty progress update), then continue with normal dispatch below.

If first token equals executables (i.e. executables alone or executables <path>, NOT a path or word that merely starts with the string executables): skip Steps 2–5 entirely and go to "Mode: Executables Extraction" below.

If first token equals prune: skip Steps 2–5 entirely and go to "Mode: Memory Pruning" below.

If first token equals memory: skip Steps 2–5 entirely and go to "Mode: Memory Distillation" below.

If first token equals external (i.e. external <source>, NOT a word that merely starts with the string external): a second token is required — it is the source to analyse. No second token: print ! MISSING — `external` needs a source: `/foundry:distill external <url-or-path>`. and stop; do not enter the mode, do not guess a source, do not fall back to default-mode analysis. Second token present: skip Steps 2–5 entirely and go to "Mode: External Distillation" below.

Otherwise, look for signals of repetitive or specialist work. First three git commands independent — run in parallel:

# timeout: 3000
# --- run these three in parallel ---
git log --oneline -50

git log --name-only --pretty="" -30 | sort | uniq -c | sort -rn | head -20

git log --oneline -100 | cut -d' ' -f2 | sort | uniq -c | sort -rn | head -15

Then use Glob tool (pattern todo_*.md, path .plans/active/) to list active task files; read each with Read tool. Also read .notes/lessons.md (if exists) for task history and conversation hints.

If $ARGUMENTS provided, use as additional context for pattern analysis.

Frequency Heuristics

• 3+ occurrences of pattern in recent history: candidate for automation
• 2+ different projects using same manual process: cross-project skill
• significant manual effort per occurrence (subjective — use git history context): high-value automation target
• Domain-specific knowledge required: candidate for specialist agent (not just skill)

With --eager (lower thresholds):

• 2+ occurrences: candidate for automation
• 1 occurrence with significant manual effort: qualifies as high-value candidate
• Domain-specific threshold unchanged

export CSID="${CLAUDE_CODE_SESSION_ID:-$PPID}"
IFS= read -r _KEEP < "${TMPDIR:-/tmp}/distill-state-${CSID}/keep-items" 2>/dev/null || _KEEP=""
_PRESERVE="run-dir=n/a"
[ -n "$_KEEP" ] && _PRESERVE="$_PRESERVE; user-keep: $_KEEP"
python "${CLAUDE_PLUGIN_ROOT:-plugins/cc_foundry}/bin/write_skill_contract.py" "foundry:distill" "gap-analysis (after work-pattern scan)" "n/a" "$_PRESERVE" "gap analysis (Step 3) → duplication check (Step 4) → report (Step 5)"  # timeout: 5000

Step 3: Gap analysis

For each identified pattern, check:

1. Already covered? — search existing agent/skill descriptions for overlap
2. Frequent enough? — recurring ≥3 times or clearly domain-specialized (see Step 2 heuristics — combine ≥3 occurrences with effort/frequency signals from Steps 1–2)
3. Would specialist add quality? — needs deep domain knowledge?
4. Too narrow? — single-use task doesn't warrant persistent agent

Thresholds for recommendation:

• New agent: recurring specialist role, complex decision-making, 5+ distinct capabilities
• New skill: workflow orchestration, multi-step process with fixed structure
• No new file needed: one-off or already covered by existing agent

Step 4: Check for duplication

Before recommending anything, run overlap check and anti-pattern checklist:

For each candidate agent/skill:
- Does any existing agent cover >50% of its scope? → enhance existing instead
  (with --eager: lower to >30%; any shared single named capability → flag as boundary issue)
- Is the name/description confusingly similar to an existing one? → rename existing

Anti-pattern checklist — reject candidate if any apply:

1. Role vs task confusion: agents are roles, not tasks. Don't create agent per different topic.
2. Near-duplicate: candidate duplicates existing agent, slightly different name. Enhance existing instead.
3. Thin wrapper: candidate skill just calls one agent with fixed args. Not enough value for new skill file. Exception: skills adding measure-first/measure-after bookends, multi-mode dispatch across 3+ agents, or safety breaks (retry limits, validation gates) justify wrapper even if only one agent executes for given invocation.

Step 5: Report

## Agent/Skill Suggestions

### Existing Coverage (no gaps found)
- [agent/skill]: covers [pattern] well — no new file needed

### Recommend: New Agent — [name]
**Trigger**: [what recurring pattern or gap justifies this]
**Gap**: [what existing agents don't cover]
**Scope**: [what it would do — 3-5 bullet points]
**Suggested tools**: [Read, Write, Edit, Bash, etc.]
**Draft description**: "[one-line description for frontmatter]"

### Recommend: New Skill — [name]
**Trigger**: [what repetitive workflow justifies this]
**Gap**: [why existing skills don't cover it]
**Scope**: [what workflow steps it would orchestrate]
**Draft description**: "[one-line description for frontmatter]"

### Recommend: Enhance Existing — [agent/skill name]
**Add**: [specific capability missing from current version]
**Why**: [what recurring task would benefit]

### No Action Needed
[pattern]: already handled by [existing agent/skill]

## Confidence
**Score**: [0.NN]
**Gaps**:
- (-0.NN) [e.g., git history too shallow, task files not present, descriptions too generic to compare]

**Refinements**: N passes. [Pass 1: <what improved>. Pass 2: <what improved>.] — omit if 0 passes

rm -f .temp/state/skill-contract.md  # clear contract — skill complete (compaction-contract.md §Lifecycle)  # timeout: 5000

Mode: Memory Pruning — only when $ARGUMENTS == "prune"

DISTILL_MODES=$(python "${CLAUDE_PLUGIN_ROOT:-plugins/cc_foundry}/bin/resolve_skill_subdir.py" distill modes 2>/dev/null || echo "plugins/cc_foundry/skills/distill/modes")  # timeout: 5000
cat "$DISTILL_MODES/prune.md"

Execute the mode loaded above.

Mode: Memory Distillation — only when first token is memory

DISTILL_MODES=$(python "${CLAUDE_PLUGIN_ROOT:-plugins/cc_foundry}/bin/resolve_skill_subdir.py" distill modes 2>/dev/null || echo "plugins/cc_foundry/skills/distill/modes")  # timeout: 5000
cat "$DISTILL_MODES/memory.md"

Execute the mode loaded above.

Mode: External Distillation — only when $ARGUMENTS begins with external

DISTILL_MODES=$(python "${CLAUDE_PLUGIN_ROOT:-plugins/cc_foundry}/bin/resolve_skill_subdir.py" distill modes 2>/dev/null || echo "plugins/cc_foundry/skills/distill/modes")  # timeout: 5000
cat "$DISTILL_MODES/external.md"

Execute the mode loaded above.

Mode: Executables Extraction — only when $ARGUMENTS begins with executables

DISTILL_MODES=$(python "${CLAUDE_PLUGIN_ROOT:-plugins/cc_foundry}/bin/resolve_skill_subdir.py" distill modes 2>/dev/null || echo "plugins/cc_foundry/skills/distill/modes")  # timeout: 5000
cat "$DISTILL_MODES/executables.md"

Execute the mode loaded above.

</workflow> <notes>

• Skill is introspective: looks at tooling itself, not just code

• Invoke periodically (e.g. monthly) or after burst of correction/feedback; one-time snapshot, not continuous monitor

• Suggestions are proposals — review before creating new files

• After creating new agent/skill from suggestion, re-run skill once to confirm gap resolved, then stop

• memory mode is primary consolidation path — run after any session with significant corrections, prevents lesson drift into MEMORY.md noise

• Agent Teams signal tracking: when reviewing patterns, also look for:

  • Skills using --team or team-mode heuristics more/less than expected: flag over/under-use relative to decision matrix in CLAUDE.md § Agent Teams
  • Security findings in reviews for non-auth code: foundry:qa-specialist teammate scope too broad; narrow it
  • Model tier mismatches (e.g. heavy analysis assigned to sonnet teammates): flag for tier adjustment

• external mode calibration: two concrete GT fixture cases defined in calibrate skills mode file — find via find "${CLAUDE_PLUGIN_ROOT:-plugins/cc_foundry}" -maxdepth 5 -path "*/calibrate/modes/skills.md" 2>/dev/null | head -1 with fallback to plugins/cc_foundry/skills/calibrate/modes/skills.md:

  • caveman plugin — narrow, self-contained communication mode, no local structural overlap: GT install-as-is recommended, Group A empty or thin
  • Karpathy autoresearch — research automation tool, strong overlap with research: plugin structure: GT Group A candidates map to research plugin, digest recommended, install-as-is not triggered
  • Ground truth = static snapshot of each tool's agent/skill/rule files (no live fetch needed); score adoption-table lane assignments against GT outcomes

• Follow-up chains:

  • Suggestion accepted for new agent/skill: /foundry:manage create to scaffold and register it
  • Suggestion to enhance existing: edit agent/skill directly, then /foundry:setup
  • memory proposals applied: /foundry:setup to propagate; /foundry:audit rules to verify new rule files structurally sound
  • executables extraction complete: /foundry:setup to propagate bin/ scripts; run /foundry:audit --efficiency to confirm clusters == 0

</notes>