agentleFS
samwelmungasamwelmunga/jenga-npm/.agents/skills/j-smoke-test/SKILL.md

j.smoke-test

Skill0 starsChanged 23 days agoApache-2.0View on GitHub

Run a structural smoke test across every skill in the catalog (frontmatter, name versus directory, preferred agent, referenced scripts and assets) and report a pass/fail summary with findings classified as critical or minor.

The file

We split this skill into the rules it gives an agent

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

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    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-only: it never edits a skill and never writes board status.

    line 25
  2. Distinctive:

    Never fix a skill unprompted: the user chooses what to change.

    line 68
  3. Very specific:

    When the user is done, remove $FINDINGS.

    line 70

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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

Smoke Test — Does every skill still hang together?

skills/j-smoke-test/ is the canonical, hand-edited directory for this skill; the frontmatter name is j.smoke-test, never j.j-smoke-test. This file only sequences the harness and interprets its result. Every check, the critical-versus-minor severity bar and the output format live in skills/j-smoke-test/scripts/smoke-test-skills.sh, whose header is the contract; the critical-error table is rendered by skills/j-smoke-test/scripts/render-error-table.sh, likewise. Do not restate or re-implement them here.

Boundaries

  • Structural, not functional. It checks that each skill loads and its references resolve. It does not run tests/*.bats, and a clean smoke test says nothing about whether a skill behaves correctly.
  • It is not scripts/smoke-harness.sh, which smoke-tests the diff of one in-flight inline task.
  • Read-only: it never edits a skill and never writes board status. A finding is information for the user, not a verdict on any board item.

Instructions

Resolve the scripts directory once, and create one scratch file for the findings:

SMOKE_DIR="$([ -d skills/j-smoke-test/scripts ] && echo skills/j-smoke-test/scripts || echo node_modules/@jenga-ai/agent/skills/j-smoke-test/scripts)"
FINDINGS="$(mktemp)"

Shell variables do not survive between separate commands: define SMOKE_DIR again in each later command, and use the literal findings path that mktemp printed.

1. Run the harness

bash "$SMOKE_DIR/smoke-test-skills.sh" > "$FINDINGS"

Its stderr is the pass/fail summary: relay it unchanged. Findings are JSON lines in $FINDINGS. Exit 0: no critical finding. Exit 1: at least one critical finding. Exit 2: usage error or jq missing; show its stderr and stop.

2. Render the critical-error table

bash "$SMOKE_DIR/render-error-table.sh" "$FINDINGS"

Print its stdout verbatim. The table (columns Feature, Error, Consequences, shaped by templates/CRITICAL_ERROR_TABLE_TEMPLATE.md and registered as the error_table type in templates/playbook-types.json) holds only the critical findings. When there is no critical finding the script prints nothing: print no table, no heading and no placeholder for it. Exit 2: show its stderr and stop.

3. Interpret the result

  • Exit 0 with no findings: say every skill passed.
  • Exit 0 with minor findings: say the catalog passed and list the minor findings briefly (skill, error).
  • Exit 1: the table from step 2 is the report; do not restate its rows. Offer next steps:

    What would you like to do?
    1. Fix one of the findings now (name it)
    2. Capture the findings as board items with j.todo
    3. Stop here
    4. Other (describe below)

Never fix a skill unprompted: the user chooses what to change.

When the user is done, remove $FINDINGS.

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.