swarm

snappy-swarm/api.ts - How to construct agent prompts for BOTH code-review and browser-testing agents, collect results, resolve conflicts, and decide when to stop.

Ported from kernel snappy-swarm in Phase 0.5. See state/lib/swarm.ts for the full API surface.

Steps

• See state/lib/swarm.ts

Eval

Actor: the exported functions in state/lib/swarm.ts. Auditor: none wired yet - eval is manual (Robert review). File a state/log/pending-eval.ndjson row on each run.

Score convention:

Gotchas

via the Phase 0.5 driver. Only these rewrites were applied: already in state/lib/)

1. realpathSync(process.argv[1]) CLI guard wrapped in try/catch

• See the kernel SKILL.md for the original long-form guidance if you need it

(read-only reference at the kernel path above).

Graduation

This skill is prose. Graduate by defining a deterministic auditor and flipping eval: auto.

Rubric

criteria:
  - name: api_exists
    kind: deterministic
    check: "The file 'state/lib/swarm.ts' should exist and export functions for prompt construction, result collection, conflict resolution, and iteration stopping."
  - name: log_entry_created
    kind: deterministic
    check: "A new entry should be appended to 'state/log/pending-eval.ndjson' after the skill execution."
  - name: conflicts_resolved
    kind: judge
    check: "The output of the skill should demonstrate a clear strategy for resolving conflicts between subagent results and a decision on when to stop iterating."

swarm - loader

Per-turn rules for the swarm skill. Full reference: state/skills/swarm/SKILL.md.

Critical Rules

_(no failures recorded yet - this skill is a Phase 0.5 mechanical port from kernel snappy-swarm and has not produced hard-won rules in mini. Read state/skills/swarm/SKILL.md and state/lib/swarm.ts before invoking.)_

• The lib is a scaffold stub per the index; treat any production use as exploratory.
• Eval is manual (Robert review) - every run MUST file a row in state/log/pending-eval.ndjson.

Commands

| ui dashboard | state/skills/swarm/resources/ui.openui | |invoke: import from state/lib/swarm.ts |verify: npx tsx state/lib/swarm.ts (import-check) |eval log: state/log/pending-eval.ndjson (skill: "swarm" - manual review queue)

OpenUI Resource

• Skill-owned OpenUI Lang resource: state/skills/swarm/resources/ui.openui. Read it before rendering or editing this skill's generated component surface.
• Treat this resource as a first-class artifact of the skill, not a generic chat response. Improve it when the skill's user-facing output needs to become richer.
• System resources compose OpenUI primitives and inherit SnappyChat tokens. Use ui_contract: branded in SKILL.md only for deliberate platform or client visuals.

Known Pitfalls

• Phase 0.5 ported file: only mechanical rewrites applied (path patches, realpathSync CLI guard wrapped in try/catch). Original kernel SKILL.md at ~/.claude/skills/snappy-swarm/ is the long-form reference.
• No deterministic auditor exists - graduation requires defining one and flipping eval: auto.

Self-Test

An agent reading this should correctly:

1. [ ] File a pending-eval.ndjson row instead of an evals.ndjson row
2. [ ] Read state/lib/swarm.ts before assuming behavior
3. [ ] Treat the skill as scaffold, not graduated

Self-report

If this loader fell short, append a line:

echo "[$(date -u +%FT%TZ)] swarm: <what was missing>" >> state/log/loader-feedback.log

<!-- kernel-ok: historical attribution for the PID-loader pattern. -->

Self-correcting loader (PID feedback)

This is the rule that makes loaders learn. Ported from snappy-kernel, where it demonstrably worked: the snappy-dom-cartographer skill mapped Skool admin UI once, every later ship-agent inherited the full map for free.

Before you finish, do two things:

1. Fix gaps (P - proportional). MANDATORY. If this loader didn't cover your case - if you had to read another file, run an undocumented command, work around a wrong selector, or discover a quirk - you MUST attempt an Edit to this AGENTS.md before you log.

• Default action: edit this file. Surgical. Add the missing line, table

row, or code example where it belongs. No restructuring.

• Only LOGGED is allowed when: the fix needs >10 lines, spans multiple

files, or requires a structural rewrite. In that case the state/regen/drain.sh queue picks it up asynchronously.

• "I didn't have time" / "it's minor" / "the next agent will figure it out"

are NOT valid reasons. Minor gaps compound - that's the whole point of this loop.

• Concrete example of a valid P-fix: discover agent-browser screenshot

takes path positionally not as --output <path>. One line in a table. Edit it in. Do not just log it.

• The goal: the next agent never has to leave the loader.

2. Log the result.

echo "[$(date -u +%FT%TZ)] swarm: <what was missing or fixed> [FIXED|LOGGED]" >> state/log/loader-feedback.log

• FIXED = you patched this loader inline (P-fix).
• LOGGED = too large for inline; the PostToolUse enqueue + Stop-hook drain

will rewrite the loader from scratch on next session-end.

Do not skip this. Every agent run must leave the system better than it found it. The loader is the setpoint; you are the sensor; the gap is the error signal; closing the gap is the correction.