Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/fortunto2/rust-code/swarm-improvenpx skills add fortunto2/rust-code --skill swarm-improvegit clone --depth 1 https://github.com/fortunto2/rust-codeWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00089 | $0.01396 |
| Opus 5 | $0.00044 | $0.00698 |
| Sonnet 5 | $0.00018 | $0.00279 |
| Haiku 4.5 | $0.00009 | $0.00140 |
Grade A, and why
swarm-improve scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 115 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/swarm-improve — Multi-Agent Project Improvement
End-to-end workflow: plan → branch → delegate execution → verify.
This is a thin orchestrator. The heavy lifting is in /plan (analysis) and /delegate (execution).
Step 0: Validate Baseline
Before anything else:
cdto the project (absolute path — store it, you'll need it for every delegate)- Read
CLAUDE.mdorREADME.md→ find the test command - Run tests:
bash: make test(or equivalent)
If tests FAIL → HARD STOP. Create one P0 task "fix broken tests" and finish. Do NOT plan improvements on a broken project — all delegate work will fail too.
Step 1: Plan
If .tasks/PLAN.md already exists and is recent (< 24h):
- Read it, skip to Step 2
- Tell the user you're using the existing plan
Otherwise, run the /plan workflow inline (this skill has the same tools):
- Recon — already done in Step 0
- Design dimensions — pick 3+ analysis angles adapted to the goal (not hardcoded)
- Check agent availability — try each agent before committing to the plan
- Delegate to 3+ diverse agents — agent diversity mandatory (claude + gemini + codex preferred)
- Wait — poll every 60s with
delegate_status - Synthesize — cross-reference, dedup, create STAR-formatted execution tasks +
.tasks/PLAN.md
Use the task tool for all task creation — never write .tasks/ files manually. Files are named YYYYMMDD-NNN-slug.md automatically.
Step 2: Branch
bash: cd <absolute-project-path> && git checkout -b improve/swarm-$(date +%Y%m%d)
All delegate work happens on this branch. Master/main stays clean.
Step 3: Delegate Execution
Read .tasks/PLAN.md for the prioritized task list. Each task has STAR format — Action tells the delegate what to do, Result tells how to verify.
Assign agents by strength, diversify across agents:
| Task type | Primary | Fallback |
|---|---|---|
| Code changes, refactoring | claude | gemini |
| Quick fixes, single-file | codex | claude |
| Tests, docs | claude | codex |
| Review, analysis | gemini | opencode |
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 115 lines · 89 tokens per session scan A dcf96e1ee5d7
swarm-improve is a skill published in the GitHub repository fortunto2/rust-code (41 stars, last pushed 10d ago), licensed MIT. It adds 89 tokens to every session and 1,396 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
test-repo
Use this skill to test strategy changes against a fresh test repository. Invoke when the user asks to "test against a test repo", "validate the changes", or wants to verify session hooks, commits, and checkpoint creation work correctly.
development
开发语言能力索引。Python、Go、Rust、TypeScript、Java、C++、Shell。当用户提到编程、开发、代码、语言时路由到此。.
model-update
新增或更新 CCX 项目中的模型注册技能。适用于 Claude 与非 Claude 模型，覆盖能力表、路由白名单、基准映射、代码生成与验证。.
conductor-implement
Executes the tasks defined in the specified track's plan. Use this to start or continue working on a feature, bug fix, or chore.
skills-audit
Audit .agents/skills SKILL.md files. Use for recurring checks of duplicate, overlapping, stale, inconsistent, or broken skills and merge/delete candidates.
bionic-decision-engine
Unified mathematical arbitrator for all resource allocation decisions — money, time, energy, relationships. Absorbs 46 decision protocols + 24 strategy protocols into one dense engine.