threadwork

A workflow for coordinating several named roles, such as researchers, writers, coders, and reviewers, through defined multi-step patterns.

In plain words
What is it for?
Use it for research-then-writing, parallel reviews, or a code-review-and-fix cycle with recorded run traces.
Why use it?
It organizes handoffs and feedback loops for work that needs research, review, or repeated fixes.

Skill for Claude CodeCodex

Install

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.

agentmods
npx agentmods add skills/tianqi-bu/threadwork/skill
Any agent
npx skills add Tianqi-Bu/threadwork --skill skill
Clone the repo
git clone --depth 1 https://github.com/Tianqi-Bu/threadwork

Made for: Claude Code, Codex.

Per session 54 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 746 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
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 $0.00054 $0.00746
Opus 5 $0.00027 $0.00373
Sonnet 5 $0.00011 $0.00149
Haiku 4.5 $0.00005 $0.00075

Measured 2d ago against content hash 1e01e5f3e455, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

threadwork 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.

The scan reads SKILL.md. This mod also ships 1 executable file (scripts/lint-skill.mjs), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

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.

packages/skill/SKILL.md · 78 lines

How it starts

The opening of the file, as written. The whole thing — 78 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Threadwork

Threadwork dispatches work across yaml-defined roles, records step-level traces to SQLite, and lets you replay the run as a static HTML timeline.

Patterns

This skill supports three orchestration patterns. Pick the one that matches the task; the parent session is responsible for invoking the corresponding sub-agents and recording trace events.

research-then-write

Sequential pipeline. The researcher role gathers cited facts; the writer role turns those notes into prose. Use when the task is "produce a written artefact about X" and X needs investigation first.

See patterns/research-then-write.md.

parallel-review

Three reviewers (reviewer × 3 by default) examine the same artefact in parallel and emit verdicts. Threadwork majority-votes on the first token of each verdict (APPROVED / REVISE / REJECT). Use when a piece of work needs more than one set of eyes and the cost of a missed defect is real.

See patterns/parallel-review.md.

code-review-fix

coder produces a change; reviewer evaluates. If the verdict is not APPROVED, the reviewer's notes feed back into a second coder iteration. The loop bounds at three iterations by default. Use for single-file bug fixes or short feature additions where reviewer feedback is genuinely actionable.

See patterns/code-review-fix.md.

Pre-conditions

Before invoking the skill the parent session should:

  1. Have threadwork-cli installed and on PATH.
  2. Have the Threadwork MCP server registered in ~/.claude.json (run threadwork init once; details land in W7).
  3. Have at least the four default roles installed: researcher, coder, reviewer, writer. threadwork roles list should show them.

Invocation contract

/threadwork <pattern> <task>

  • <pattern> — one of research-then-write, parallel-review, code-review-fix.
  • <task> — a free-form description of what to produce.

Read the full file on GitHub · 78 lines

Files

What ships with it

5 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. 2d ago First seen · 78 lines · 54 tokens per session scan A 1e01e5f3e455

Subscribe to this mod's changes

threadwork is a skill published in the GitHub repository Tianqi-Bu/threadwork (1 stars, last pushed 1mo ago), licensed MIT. It adds 54 tokens to every session and 746 once invoked, about $0.0003 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

git-workflow-and-versioning

Structures git workflow practices. Use when making any code change. Use when committing, branching, resolving conflicts, or when you need to organize work across multiple parallel streams.

shashankswe2020-ux/whoop-mcp · 40 tokens

security-and-hardening

Hardens code against vulnerabilities. Use when handling user input, authentication, data storage, or external integrations. Use when building any feature that accepts untrusted data, manages user sessions, or interacts with third-party services.

shashankswe2020-ux/whoop-mcp · 48 tokens

documentation-and-adrs

Records decisions and documentation. Use when making architectural decisions, changing public APIs, shipping features, or when you need to record context that future engineers and agents will need to understand the codebase.

shashankswe2020-ux/whoop-mcp · 43 tokens

coding-agents-farm

To orchestrate parallel coding-agent farms (Claude, Codex, Copilot, Gemini, etc.) on isolated git worktrees.

griddynamics/rosetta · 32 tokens

orch

AI agent orchestrator — manage teams of AI agents that work on your codebase in parallel. Use when the user wants to: run multiple agents, coordinate AI work, deploy agent teams, manage tasks/goals/agents, check orchestrator status, or mentions 'orch', 'orchestry', 'agents team', 'agent orchestration'.

oxgeneral/ORCH · 71 tokens

verify-frontend-change

Verify a frontend change actually works in the GSD Task Manager app before trusting it. Drives the running app browser-first — seeds IndexedDB, busts the PWA service-worker cache, then observes real behavior plus console/network errors — and codifies regression-worthy behavior as a Playwright spec. Use whenever you've…

vscarpenter/gsd-task-manager · 138 tokens