repo-initialize
Initialize or retrofit a consuming repo into the llm-os working model.
Skill by armin-haghi
Chalk Labs
Need help deploying repo-initialize?
Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/armin-haghi-repo-initialize
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: ml
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/repo-initialize
What's inside
- Skill name
repo-initialize- Description
- Initialize or retrofit a consuming repo into the llm-os working model.
- License
- MIT
- Instructions
- 143 lines in SKILL.md · view on GitHub
Agent Skills spec issues
- files: referenced but missing: templates/AGENTS.md, templates/project-spine.md, templates/current-milestone.md, templates/open-questions.md, templates/session-brief.md
Files (1)
- SKILL.md
SKILL.md (first lines)
# Repo Initialize Use this skill when a repo does not yet have the minimum `llm-os` operating surface in place, including an existing repo that has work and docs already but does not yet follow the model. ## Goal Bring a repo to the minimum working model so an agent can pick it up repo-first without depending on Notion or prior chat context. ## Start with audit, not scaffolding For an existing repo, inspect before creating anything. Check: - current entry surfaces such as `AGENTS.md`, `README.md`, `ASSISTANT.md`, or `CLAUDE.md` - current live docs such as `docs/`, `planning/`, `notes/`, `specs/`, or similar folders - whether the repo already has a durable milestone doc, session handoff, or project summary Map what already exists to the canonical artifacts: - `project-spine.md` - `current-milestone.md` - `open-questions.md` - `session-brief.md` - `review-report.md` - `project-overview.yaml` ## Decide the live project-doc surface Default to `docs/`. Keep an existing non-default location only when moving it would create more confusion than clarity right now. If keeping a non-default location: - declare it explicitly in the repo's local `AGENTS.md` - treat that location as the live project-doc surface - do not create a second parallel `docs/` surface just because it is the default ## Create only what is missing For a new or empty repo, create: 1. root `AGENTS.md` from `templates/AGENTS.md` 2. `docs/project-spine.md` from `templates/project-spine.md` 3. `docs/current-milestone.md` from `templates/current-milestone.md` 4. `docs/open-questions.md` from `templates/open-questions.md` 5. `docs/session-brief.md` from `templates/session-brief.md` 6. `docs/review-report.md` from `templates/review-report.md` 7. `project-overview.yaml` from `templates/project-overview.yaml` For an existing repo: 1. create root `AGENTS.md` if missing 2. keep and reuse existing docs when they already serve a canonical role 3. create only missing canonical artifacts 4. migrate or rewrite weak docs only when needed to remove ambiguity Do not duplicate an existing usable doc just to match the default filenames. Prefer one live surface with clear mapping over two partially overlapping ones. ## Normalize existing docs when needed If a repo already has planning or handoff docs: - treat the best current artifact as the starting source - rewrite or condense it into the canonical artifact shape when needed - preserve durable content, but remove duplicate or competing summaries Typical examples: - a roadmap or sprint note can become `current-milestone.md` - a handoff note can become `session-brief.md` - a project README or spec can seed `project-spine.md` - an issue list or TODO note can seed `open-questions.md` ## Fill the minimum required content Do not leave the repo as a pure empty scaffold or as a retrofit with only file renames and no usable content. Fill at least: - the live `project-spine.md` - what - why - who - constraints - success criteria - canonical links - the live `current-milestone.md` - snapshot - why this milestone matters - in scope - out of scope - acceptance criteria - the live `session-brief.md` - snapshot - project summary - next action - blocking question - expected output from this session - `project-overview.yaml` - project - status - stage - active_milestone - freshness - last_reviewed - owner - blocker - next_decision - next_action - canonical_repo - execution_surface - commercial_signal The live `open-questions.md` and `review-report.md` may start sparse, but they should exist. ## Repo-local overrides If the repo needs a non-default live project-doc location or any other
From armin-haghi/llm-os, MIT.
More from armin-haghi/llm-os
- 0.497score
pitch-deck-builder
armin-haghi
Build a complete, polished pitch or reference deck from scratch.
SkillCFreshness: C★ 1Claude Code +9 - 0.494score
personal-project-planner
armin-haghi
Shape a personal tool or side-project into a small, buildable plan and write back the durable outcomes into the llm-os project docs.
SkillCFreshness: C★ 1Claude Code +9 - 0.494score
business-case-analyzer
armin-haghi
Analyze a business or product idea through opportunity framing and solution shaping, then write back the durable outcomes into the llm-os project docs.
SkillCFreshness: C★ 1Claude Code +9 - 0.491score
orchestrator
armin-haghi
Use the single cross-interface orchestration path for project sweeps and selected-project agent work.
SkillCFreshness: C★ 1Claude Code +9 - 0.491score
project-audit
armin-haghi
Audit one project for LLM-OS readiness, gaps, compatibility, and source-of-truth drift.
SkillCFreshness: C★ 1Claude Code +9 - 0.491score
morning-status-brief
armin-haghi
Produce a read-only morning project status from the LLM-OS Project Control Tower and Agent Run Queue.
SkillCFreshness: C★ 1Claude Code +9 - 0.426score
surface-human-topics
armin-haghi
Surface decision-worthy and uncertainty-heavy topics inside one project's active context.
SkillDFreshness: D★ 1Claude Code +9 - 0.426score
capture-human-input
armin-haghi
Capture lightweight human input and write it to the smallest correct durable artifact.
SkillDFreshness: D★ 1Claude Code +9 - 0.426score
project-refresh
armin-haghi
Refresh canonical project context and tighten milestone clarity.
SkillDFreshness: D★ 1Claude Code +9 - 0.426score
review-gate
armin-haghi
Review milestone output and produce a durable pass / conditional pass / fail result.
SkillDFreshness: D★ 1Claude Code +9 - 0.422score
agent-run
armin-haghi
Pick up and advance one bounded run from the Agent Run Queue.
SkillDFreshness: D★ 1Claude Code +9 - 0.422score
llm-os-session
armin-haghi
Start or continue serious work using the canonical llm-os core.
SkillDFreshness: D★ 1Claude Code +9
Signals
- Stars
- 1
- Forks
- 0
- Open issues
- 1
- Last commit (this folder)
- Apr 25, 2026 (6 months ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/armin-haghi/llm-os
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from armin-haghi
- 0.497score
pitch-deck-builder
armin-haghi
Build a complete, polished pitch or reference deck from scratch.
SkillCFreshness: C★ 1Claude Code +9 - 0.494score
personal-project-planner
armin-haghi
Shape a personal tool or side-project into a small, buildable plan and write back the durable outcomes into the llm-os project docs.
SkillCFreshness: C★ 1Claude Code +9 - 0.494score
business-case-analyzer
armin-haghi
Analyze a business or product idea through opportunity framing and solution shaping, then write back the durable outcomes into the llm-os project docs.
SkillCFreshness: C★ 1Claude Code +9 - 0.491score
orchestrator
armin-haghi
Use the single cross-interface orchestration path for project sweeps and selected-project agent work.
SkillCFreshness: C★ 1Claude Code +9 - 0.491score
morning-status-brief
armin-haghi
Produce a read-only morning project status from the LLM-OS Project Control Tower and Agent Run Queue.
SkillCFreshness: C★ 1Claude Code +9 - 0.491score
project-audit
armin-haghi
Audit one project for LLM-OS readiness, gaps, compatibility, and source-of-truth drift.
SkillCFreshness: C★ 1Claude Code +9
Done for you
Rather not set it up yourself?
Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.
See how we build AI agents →Read the case studies
Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.
- launches shipped
- 50+
- launches shipped
- media placements
- 500+
- media placements
- founders on every account
- 2
- founders on every account
Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block
Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.
Get repo-initialize running in your business
Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.
Cite this page
repo-initialize skill. Agent Mart. https://agentmart.dev/skills/armin-haghi-repo-initialize (last updated 2026-10-24)