agentleFS
tanishq286tanishq286/tq-forge/skills/tq-forge-improve-all/SKILL.md

tq-forge-improve-all

Skill8 starsChanged 5 months agoMITView on GitHub

Scan every forged skill and agent for scores below 7, then improve each one in priority order (lowest first). Runs assess -> rewrite -> verify per artifact, stopping if the halt flag appears or the user interrupts. Use when asked for "/tq-forge-improve-all", "improve everything", "fix all weak skills", or "bulk improve".

The file

We split this skill into the rules it gives an agent

2 rules. About 0 are common, 1 is distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Read the artifact, rewrite only the weak sections, then:

    line 64
  2. Very specific:

    If an agent's tools.json is malformed, skip that agent with a warning rather than aborting the whole batch.

    line 86

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

/tq-forge-improve-all — batch improve every sub-7 artifact

When to use

You want a single sweep that finds every forged skill and agent scoring below 7 and fixes them in one session. Good after an initial forge spree, or after /tq-forge-scan surfaces several weak entries. Each artifact gets one improve attempt; if it still fails, it's flagged for manual review and the loop moves on.

Procedure

export TQ_FORGE_HOME="${TQ_FORGE_HOME:-$HOME/.tq-forge}"
S="${CLAUDE_PLUGIN_ROOT:-${TQ_FORGE_HOME:-$HOME/.tq-forge}/install}/scripts"
source "$S/common.sh" && tq_ensure_home
  1. Enumerate artifacts below threshold. From the log plus any un-logged sandbox dirs:
    python3 - <<'PY'
    import json, pathlib, os
    home = pathlib.Path(os.environ['TQ_FORGE_HOME'])
    log = json.loads((home/'skill-log.json').read_text() or '[]')
    weak = [e for e in log if isinstance(e.get('score'),(int,float)) and e['score'] < 7]
    weak.sort(key=lambda e: e['score'])
    for e in weak:
        print(f"{e['slug']}\t{e['score']}\t{e.get('status','?')}")
    PY
  2. Print the improvement queue so the user can abort if it's longer than expected:
    🔧 Improvement queue — <N> artifacts below 7/10
     1. notion-daily-summary    5.8/10  (weak: actionability, specificity)
     2. inventory-snapshot      6.2/10  (weak: completeness)
  3. For each artifact in order: a. Check the halt flag before each item:
    test -f "$TQ_FORGE_HOME/halt.flag" && echo "⏸  halt.flag set — stopping batch." && exit 0
    b. Run the /tq-forge-improve logic inline:
    bash "$S/improve-loop.sh" assess "<slug>"
    Read the artifact, rewrite only the weak sections, then:
    bash "$S/improve-loop.sh" verify "<slug>"
    c. Print a one-line status: ✅ notion-daily-summary: 5.8 -> 8.1 or ⚠️ inventory-snapshot: 6.2 -> 6.5 — still weak, flagged for review.
  4. Print the final summary.
    📊 Improve-all complete
    Improved:   <N>  (now >=7)
    Still weak: <M>  (flagged in forge-queue.json)
    Skipped:    <K>  (halt.flag hit mid-run)
  5. If any remain in needs_manual_review, print: "Run /tq-forge-queue to see what needs a human touch."

Pitfalls

  • An artifact scoring 6.9 on one run can score 7.1 on the next with no edit (scorer variance). Accept >=7.0 as passing; don't re-run needlessly.
  • If an agent's tools.json is malformed, skip that agent with a warning rather than aborting the whole batch.
  • Promoted artifacts under $CLAUDE_SKILLS_DIR are editable here — the path is already writable. Don't treat "promoted" as read-only.

Verification

  • skill-log.json score fields are updated for every processed artifact.
  • The printed "Improved + Still weak + Skipped = N" matches the step-2 queue.
  • Running the skill again immediately prints "0 artifacts below 7" if all passed.

Tags

tq-forge improve batch quality bulk

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.