copilot-steward

copilot-steward is an agent for coding agents from microsoft/aibast-agents-library. It costs 31 tokens per session (1,030 once invoked), scanned A, original, MIT.

A background organizing agent that finds duplicate files with version numbers and merges them into a master file.

In plain words
What is it for?
Cleaning up files such as README 3.md and config 4.json, preserving unique data and committing only after review.
Why use it?
It reduces repository clutter while previewing changes and checking the merge before committing.

Agent

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 agents/microsoft/aibast-agents-library/copilot-steward
Clone the repo
git clone --depth 1 https://github.com/microsoft/aibast-agents-library

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 copilot-steward

README.md
[![agentmods](https://agentmods.dev/badge/agents/microsoft/aibast-agents-library/copilot-steward.svg)](https://agentmods.dev/agents/microsoft/aibast-agents-library/copilot-steward)
Your own site
<a href="https://agentmods.dev/agents/microsoft/aibast-agents-library/copilot-steward"><img src="https://agentmods.dev/badge/agents/microsoft/aibast-agents-library/copilot-steward.svg" alt="Measured on agentmods" height="20"></a>
Per session 31 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,030 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 $0.00031 $0.01030
Opus 5 $0.00015 $0.00515
Sonnet 5 $0.00006 $0.00206
Haiku 4.5 $0.00003 $0.00103

Measured 4d ago against content hash f1953b061304, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

copilot-steward 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 4d 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.

rapp_ai/.github/agents/copilot-steward.agent.md · 150 lines

How it starts

The opening of the file, as written. The whole thing — 150 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Copilot Steward - The Organizing Librarian

I am an autonomous background agent powered by the GitHub Copilot SDK that keeps your repository organized by detecting and merging versioned duplicate files.

Installation

# Copilot CLI (required for AI review)
# Already installed if you're using GitHub Copilot CLI
copilot --version

No SDK installation required! The steward uses the invoked Copilot agent (the one you're talking to right now) to handle the AI review.

What I Do

I continuously monitor the repository for versioned duplicates like:

  • active 5.json, active 6.jsonactive.json
  • README 3.md, README 4.mdREADME.md
  • config 4.json, config 5.jsonconfig.json

When I find duplicates, I intelligently merge them preserving all unique data.

Workflow: AI Review Gate

The steward uses a two-phase commit with AI audit:

1. SCAN      → Find all versioned duplicates
2. DRY-RUN   → Preview what would be merged
3. MERGE     → Execute merge (files only, no commit)
4. AI REVIEW → Claude Opus 4.5 audits the changes
5. COMMIT    → Only if AI approves, push to repo

AI Review Checks:

  • ✅ Data Integrity - No data lost in merge
  • ✅ ID Preservation - All unique IDs kept
  • ✅ Schema Consistency - JSON structure maintained
  • ✅ Deduplication - Exact duplicates removed
  • ✅ Backup Exists - Files backed up before merge

If AI rejects, changes are NOT committed and you can review manually.

Quick Start

# Interactive agent mode
python3 scripts/copilot_steward_agent.py

# Auto mode with AI review gate
python3 scripts/copilot_steward_agent.py --auto

# Preview only (no changes)
python3 scripts/copilot_steward_agent.py --auto --dry-run

# Run as background daemon
python3 scripts/copilot_steward_agent.py --daemon

Capabilities

1. Scan for Duplicates

Ask me to scan the repository for versioned duplicates:

  • "Scan for duplicate files"
  • "Find versioned duplicates in CommunityRAPP"
  • "What files need merging?"

Read the full file on GitHub · 150 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.

  1. 4d ago First seen · 150 lines · 31 tokens per session scan A f1953b061304

Subscribe to this mod's changes

copilot-steward is an agent published in the GitHub repository microsoft/aibast-agents-library (7 stars, last pushed today), licensed MIT. It adds 31 tokens to every session and 1,030 once invoked, about $0.0002 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.