swarm-mcp: Skill for Claude Code

.agents/skills/swarm-deepdive/SKILL.md

swarm-deepdive is a skill for Claude Code, Codex from Volpestyle/swarm-mcp. It costs 88 tokens per session (2,143 once invoked), scanned A, original, MIT.

A forensic investigation tool for swarm-based coding work, where several agents coordinate on tasks. It examines stored activity and logs to reconstruct what happened.

In plain words
What is it for?
Use it to rebuild timelines, inspect task and message history, audit file locks or shared values, investigate pending UI commands, and write postmortems.
Why use it?
It helps explain stuck agents, failed tasks, missing messages, file conflicts, and other coordination problems using recorded evidence.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one. Also seen: installed under .agents/ (shared by several agents).

This is Volpestyle/swarm-mcp's own configuration. It tells Claude Code and Codex how to work on swarm-mcp itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything swarm-mcp configures →

Reuse

Borrowing it

Nothing to install: this file belongs to Volpestyle/swarm-mcp. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/Volpestyle/swarm-mcp/main/.agents/skills/swarm-deepdive/SKILL.md
Clone the repo
git clone --depth 1 https://github.com/Volpestyle/swarm-mcp

Made for: Claude Code, Codex.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for swarm-deepdive

README.md
[![agentmods](https://agentmods.dev/badge/skills/volpestyle/swarm-mcp/swarm-deepdive/github.svg)](https://agentmods.dev/skills/volpestyle/swarm-mcp/swarm-deepdive)
Your own site
<a href="https://agentmods.dev/skills/volpestyle/swarm-mcp/swarm-deepdive"><img src="https://agentmods.dev/badge/skills/volpestyle/swarm-mcp/swarm-deepdive/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for swarm-deepdive

Your own site · 80×15
<a href="https://agentmods.dev/skills/volpestyle/swarm-mcp/swarm-deepdive"><img src="https://agentmods.dev/badge/skills/volpestyle/swarm-mcp/swarm-deepdive.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 88 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 2,143 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

What 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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00088 $0.02143
Opus 5 $0.00044 $0.01071
Sonnet 5 $0.00018 $0.00429
Haiku 4.5 $0.00009 $0.00214

Measured 10d ago against content hash 2d91bb91e718, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-10, from the pricing page.

Security

Grade A, and why

swarm-deepdive 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 10d 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.

.agents/skills/swarm-deepdive/SKILL.md · 159 lines

How it starts

The opening of the file, as written. The whole thing — 159 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Swarm Deep Dive

Use this skill to investigate what a swarm actually did, either live or after the fact. Normal swarm MCP tools (list_instances, list_tasks, poll_messages, kv_get) are for participation from one registered session. This skill is for inspection: read the SQLite database, event audit log, and server logs to reconstruct timelines and explain coordination failures.

If the user only wants to coordinate current work, use swarm-mcp instead. If they want evidence, timelines, root cause, or a postmortem, use this skill.

When To Use

  • "What happened in the swarm overnight / while I was away?"
  • "Why is task X blocked / why did agent Y stop?"
  • "Reconstruct the conversation between planner and implementer"
  • "Audit who edited or locked a file"
  • "Did this KV value get clobbered?"
  • "Why did this UI command stay pending?"
  • "Why didn't my agent register / why was it pruned?"
  • Any postmortem on swarm behavior.

Where The Data Lives

Source Path What it has
Shared SQLite DB ${SWARM_DB_PATH:-~/.swarm-mcp/swarm.db} Instances, tasks, messages, KV, file locks, events audit log, UI command queue
Server auth DB directory of swarm DB + /server/server.db Devices, bearer tokens, pairing sessions, server audit events - not swarm coordination state
Server logs directory of swarm DB + /server/logs/swarm-server.log.YYYY-MM-DD PTY lifecycle, pairing, HTTP/WSS/UDS errors, daemon startup/crashes

The Bun MCP server, swarm-ui, and swarm-server share the same swarm DB path. The default is ~/.swarm-mcp/swarm.db, but SWARM_DB_PATH overrides it.

Preserve Evidence First

For strict postmortems, start with sqlite3 -readonly before using swarm-mcp inspect. The CLI is convenient, but every state subcommand imports the runtime and runs cleanup before it reads. That can release tasks, delete offline instance rows, clear locks/messages for those instances, delete old TTL rows, and remove orphaned instance-scoped KV.

Read the full file on GitHub · 159 lines

Files

What ships with it

3 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

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.

  1. 10d ago First seen · 159 lines · 88 tokens per session scan A 2d91bb91e718

Subscribe to this mod's changes

swarm-deepdive is a skill published in the GitHub repository Volpestyle/swarm-mcp (1 stars, last pushed 3mo ago), licensed MIT. It adds 88 tokens to every session and 2,143 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 2026-08-31.

Related

Other skills, from other repositories

api-errors

McpError constructor, JsonRpcErrorCode reference, and error handling patterns for @cyanheads/mcp-ts-core. Use when looking up error codes, understanding where errors should be thrown vs. caught, or using ErrorHandler.tryCatch in services.

cyanheads/wikidata-mcp-server · 54 tokens

api-linter

MCP definition linter rules reference. Use when bun run lint:mcp or bun run devcheck reports a lint error or warning (format-parity, schema-is-object, name-format, server-json-, etc.) and you need to understand the rule, its severity, and how to fix it. Every rule ID the linter emits has an entry in this doc.

cyanheads/nonprofit-explorer-mcp-server · 86 tokens

api-linter

MCP definition linter rules reference. Use when bun run lint:mcp or bun run devcheck reports a lint error or warning (format-parity, schema-is-object, name-format, server-json-, etc.) and you need to understand the rule, its severity, and how to fix it. Every rule ID the linter emits has an entry in this doc.

cyanheads/wikipedia-mcp-server · 86 tokens

report-issue-framework

File a bug or feature request against @cyanheads/mcp-ts-core when you hit a framework issue. Use when a builder, utility, context method, or config behaves contrary to the documented API — not for server-specific application bugs.

cyanheads/wikidata-mcp-server · 52 tokens

report-issue-local

File a bug or feature request against this MCP server's own repo. Use for server-specific issues — tool logic, service integrations, config problems, or domain bugs that aren't caused by the framework.

cyanheads/wikidata-mcp-server · 44 tokens

api-errors

McpError constructor, JsonRpcErrorCode reference, and error handling patterns for @cyanheads/mcp-ts-core. Use when looking up error codes, understanding where errors should be thrown vs. caught, or using ErrorHandler.tryCatch in services.

cyanheads/nonprofit-explorer-mcp-server · 54 tokens