What is paperclip-mcp AGENTS.md?

Instructions for agents working with Paperclip, a system that coordinates software-development agents and their tasks.

Following Paperclip's agent protocol, handling @-mentions, managing feature branches, and coordinating engineering, QA, DevOps, and technical-writing work.

They define who can assign work, run scheduled heartbeats, review code, merge branches, and move issues through the workflow.

paperclip-mcp AGENTS.md — Instructions file for Codex and OpenCode by bruhsb: Instructions for agents working with Paperclip, a system that…

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 instructions/bruhsb/paperclip-mcp/agents-md

Clone the repo

git clone --depth 1 https://github.com/bruhsb/paperclip-mcp

Made for: Codex, OpenCode.

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.

README.md

[![agentmods](https://agentmods.dev/badge/instructions/bruhsb/paperclip-mcp/agents-md.svg)](https://agentmods.dev/instructions/bruhsb/paperclip-mcp/agents-md)

Your own site

<a href="https://agentmods.dev/instructions/bruhsb/paperclip-mcp/agents-md"><img src="https://agentmods.dev/badge/instructions/bruhsb/paperclip-mcp/agents-md.svg" alt="Measured on agentmods" height="20"></a>

Per session 3,944 This file is loaded in full into every session.

When invoked 3,944 The same file — it is already loaded in full.

Security scan B 2 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.

Model	Per session	Once invoked
Fable 5.1	$0.03944	$0.03944
Opus 5	$0.01972	$0.01972
Sonnet 5	$0.00789	$0.00789
Haiku 4.5	$0.00394	$0.00394

Measured 6d ago against content hash 1d7e54823f68, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade B, and why

paperclip-mcp AGENTS.md scanned grade B with 2 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 6d 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.

Sends data to an external URLmediumData exfiltration

A POST to an outside endpoint may be telemetry or may be exfiltration; either way the mod talks to somewhere, and you should know where.

| Create issue | `curl -s -X POST -H "Content-Type: application/json" -d '{"title":"...","status":"todo","priority":"medium","projectId":"<YOUR_PROJECT_ID>","goalId":"<YOUR_GOAL_ID>","assigneeAgentId":"<YOUR_CTO_AGENT_ID

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

5. **If `paperclip_create_issue` itself is the failing tool** (or the create call fails after one retry), fall back to curl:

AGENTS.md · 192 lines

How it starts

The opening of the file, as written. The whole thing — 192 lines — stays where its author put it; the contents beside it link to each section on GitHub.

AGENTS.md

This file documents the Paperclip-orchestrated agent protocol and BMAD integration. If you are a general contributor or an end-user of this MCP server, skip this file — see CONTRIBUTING.md for dev setup and docs/README.md for user docs. This file is for agents running inside a Paperclip control plane.

Paperclip Agent Workflow

This section is for Paperclip-orchestrated agents. Human developers can skip it.

Do not invoke any /bmad-* skills. These are for human operators only. Your workflow is below.

Orchestration Model

Scrum Master is the SOLE agent with a scheduled heartbeat (every 5 min). All 7 other agents (CEO, CTO, PM, Engineer, QA, DevOps, TechWriter) have heartbeat.enabled: false and wake ONLY on @-mention.
Max 3 agents running at any time (Scrum Master + 2 IC agents).
One issue per agent — finish it completely before picking up another.
QA is the SOLE merge owner. IC agents push feature branches and set in_review; QA's APPROVE decision performs the merge, worktree cleanup, and done transition. IC agents never call git merge or git push origin main.
CEO is the SOLE agent hirer. All new specialist agents go through CEO using the paperclip-hire-agent skill, subject to board approval. CTO, Scrum Master, PM, and everyone else must NOT invoke the skill or draft hire proposals unilaterally.

@-Mention Flow

Scrum Master (heartbeat) → picks backlog → assigns → @Engineer "PAP-XX ready"
Engineer → implements → commits → push branch → set in_review → @QA "ready for review" → EXIT
QA (sole reviewer + merger) → APPROVE (squash-merge to main + delete branch + done) / REQUEST_CHANGES (todo + @role) / ESCALATE (blocked + @CTO)
Scrum Master (next heartbeat) → catches orphaned in_review → @QA · clears stale locks · closes done epics · cleans board

Agent Protocol (IC agents: Engineer / DevOps / TechWriter)

paperclip_get_me — confirm identity.
Check PAPERCLIP_TASK_ID / PAPERCLIP_WAKE_REASON — find why you woke. You were woken by an explicit @-mention (no heartbeat fires for you); identify the trigger.
Label Bootstrap. Call paperclip_list_labels once and cache the name → uuid map for the run. If any required taxonomy labels are missing (source:*, status:refined|unrefined, type:*, agent:*), call paperclip_create_label to seed them before proceeding. Full taxonomy and colors: docs/guides/issue-creation-standard.md.
paperclip_get_inbox — find your assigned issue.
paperclip_checkout_issue — claim it, passing expectedStatuses for your role so the server atomically validates the kanban column before flipping status. Engineer / DevOps / TechWriter pass ["todo"]; QA passes ["in_review"] for code review or ["todo"] for test-writing tasks. Never retry a 409 and never retry a status-mismatch rejection. If the checkout fails for either reason, post Wake mismatch: PAP-XX is in status <X>, expected [<expected>]. Not claiming. @Scrum Master — please verify assignment. on the issue, then exit cleanly. Do not mutate any other state.

Read the full file on GitHub · 192 lines

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.

6d ago First seen · 192 lines · 3,944 tokens per session scan B 1d7e54823f68

Subscribe to this mod's changes

paperclip-mcp AGENTS.md is an instructions file published in the GitHub repository bruhsb/paperclip-mcp (2 stars, last pushed 4mo ago), licensed MIT. It adds 3,944 tokens to every session, about $0.0197 per session on Opus 5. A static security scan graded it B with 2 findings (sends data to an external url, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Building agents Project management

vscode buildNext.instructions.md

Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).

microsoft/vscode · 6,785 tokens

paperclip-mcp AGENTS.md