project-agent-bootstrapper

Bootstrap governed project skills and agent permission matrices for repositories.

Updated Aug 22, 2026
One-click install
npx skills add https://github.com/xegheplimo-web/hermes-ops --skill project-agent-bootstrapper-xegheplimo-web
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: project-agent-bootstrapper
Source: https://github.com/xegheplimo-web/hermes-ops/tree/main/skills/software-development/project-agent-bootstrapper
Command: npx skills add https://github.com/xegheplimo-web/hermes-ops --skill project-agent-bootstrapper-xegheplimo-web

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve? Starting a governed coding session requires verifying repository state, ensuring the canonical review skill exists, and defining which agents may decide, review, implement, or approve. Doing this manually is error-prone and risks granting external reviewers write access or letting memory systems own runtime state. ## Core Features & Use Cases - Repository Preflight: Records branch, exact commit SHA, dirty state, and a RUN_ID for every bootstrap run, refusing to proceed outside a Git repository. - Agent Permission Matrix: Generates an agent-roles.json defining allowed and forbidden actions for 19 actors including Hermes, Devin, OpenCode, Codex, and the policy gate. - Canonical Skill Management: Validates the project-review-orchestrator skill, generates a draft if missing, and installs it only with explicit write approval. - Routing Decision: Classifies work into trivial, standard governed, or high-risk routes with recorded reasons. - Use Case: Before delegating a feature to Devin, run the bootstrapper to snapshot the repo, create the permission matrix, and produce a bootstrap report naming the next required command. ## Quick Start Ask the agent to run /project-agent-bootstrapper Prepare this repository and route agents with your repository path and objective.

Frequently Asked Questions about project-agent-bootstrapper

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I bootstrap a governed coding project with agent permissions?

Run the bootstrap_project_skill.py script with --repo and --out flags from the repository root. It verifies Git state, records the commit SHA and dirty status, writes an agent-roles.json permission matrix, and produces a bootstrap report with the next required command.

What is an agent permission matrix for AI coding agents?

It is a JSON artifact defining allowed and forbidden actions for each actor in a governed pipeline, such as orchestrator, executors, reviewers, and policy gates. It ensures external reviewers never get write access and memory systems never own runtime task state.

Does the bootstrapper work outside a Git repository?

No. The script resolves the repository root via git rev-parse and stops immediately if the path is not a Git repository. A valid commit SHA and branch are mandatory for every run.

Why does --install-if-missing do nothing on its own?

Install is nested inside draft generation, so --install-if-missing requires --generate-draft to have any effect. Without it, the script emits a WARNING action instead of silently doing nothing, and --confirm-write is also required.

When should I not use project-agent-bootstrapper?

Skip it for a single trivial typo fix, and never use it to bypass GitHub policy gates, let external reviewers modify the repository, or treat AgentMemory as authoritative runtime queue state. It prepares governance but never dispatches implementation tasks itself.