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.
npx agentmods add skills/dmythro/agent-skills/git-projectnpx skills add dmythro/agent-skills --skill git-projectgit clone --depth 1 https://github.com/dmythro/agent-skillsWrote 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.
[](https://agentmods.dev/skills/dmythro/agent-skills/git-project)<a href="https://agentmods.dev/skills/dmythro/agent-skills/git-project"><img src="https://agentmods.dev/badge/skills/dmythro/agent-skills/git-project.svg" alt="Measured on agentmods" height="20"></a>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 | $0.00176 | $0.03540 |
| Opus 5 | $0.00088 | $0.01770 |
| Sonnet 5 | $0.00035 | $0.00708 |
| Haiku 4.5 | $0.00018 | $0.00354 |
Grade A, and why
git-project 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.
How it starts
The opening of the file, as written. The whole thing — 168 lines — stays where its author put it; the contents beside it link to each section on GitHub.
GitHub Project Management
Primary skill for organizing a repository's issues into an epic-based roadmap on a GitHub Project (v2) and running the day-to-day on it. Covers setup (project, fields, views, native workflows) and operations (create epics with sub-issues, drive Status, move work, prioritize). GitHub-only, via gh project + gh api (REST sub-issues + GraphQL).
The conventions here are non-obvious and easy to get half-right: the UI nests issues by native parent/child links and tracks progress on a separate board Status field -- neither of which a markdown checklist or a closed issue touches. Get those two wrong and the work looks done while the structure silently drifts. This skill makes the flow correct-by-default.
When to Use
- Setting up a project/roadmap -- create the Project, fields (Status, Priority), the Epic + Upcoming views, enable native workflows
- Creating an epic with issues -- "make an epic X with these issues", "group these under an epic"
- Filing / moving work -- attach an issue under an epic, move an issue between epics, add new work
- Driving the board -- pick up an issue (-> In Progress), close one (-> Done), (re)prioritize
- Classifying work -- set native issue Types (Task/Bug/Feature) instead of type labels
- Milestone scoping -- create a milestone for a deliverable, assign issues, close it out; resolve "current milestone" / "milestone N"
- Migrating an existing repo's loose issues into this structure
- Configuring tool allowlists -- auto-approval patterns for read-only
gh projectcommands
Critical Rules
These are the traps -- each is a place where the obvious action leaves the structure wrong.
- Epics nest by native sub-issues, NOT markdown checklists. A
- [ ] #123bullet is cosmetic; it does not create the parent/child link the UI and roadmap read. Link natively:gh issue edit <epic> --add-sub-issue <child>/gh issue create --parent <epic>(gh >= 2.94), or the REST sub-issues API. Seereferences/sub-issues.md. - Moving an issue between epics = re-parent the native link. One command:
gh issue edit <child> --parent <newEpic>(replaces the old parent), then tidy body text. Editing bullets alone leaves it under the old epic -- the most common "looks moved but isn't" miss. - The board Status is not automatic. Closing an issue does not set
Doneunless the native Item closed workflow is enabled. Set Status explicitly, or enable the native workflows once (seereferences/project-setup.md). - Single-selects (Status, Priority) are set by option ID, not name. Look up field IDs + option IDs first (
gh project field-list <num> --owner @me --format json), thenitem-edit. Names silently no-op. - Views and workflow-enabling are one-time UI; everything else is scripted. There is no API to create/rename a view or toggle a workflow. Do those once in the UI (or copy a template); script the rest.
- Type and Milestone are issue metadata, NOT project fields. Set them on the issue (
gh issue edit <n> --type Bug --milestone "v1.0"); the board mirrors them as built-in columns for grouping/filtering, butgh project item-editcannot touch them and they never appear infield-list. Issue types exist only in org repos. Seereferences/types-and-milestones.md.
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.
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.
- 4d ago First seen · 168 lines · 176 tokens per session scan A 5ba69e827a43
git-project is a skill published in the GitHub repository dmythro/agent-skills (5 stars, last pushed 4d ago), licensed MIT. It adds 176 tokens to every session and 3,540 once invoked, about $0.0009 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.
Other skills, from other repositories
adr-writer
Writes an Architecture Decision Record (ADR) documenting a decision, its context, the options considered, and its consequences. Use to record a decision that's already been made so future readers understand why.
architecture-reviewer
Reviews application architecture for scalability, coupling, separation-of-concerns, and maintainability issues. Use for system-level review, not individual file/function-level code review.
migration-planner
Plans a safe, incremental migration (framework, database, service, or platform) with rollback points and a dual-running strategy. Use when replacing something that's already in production and can't just be swapped in one step.
project-planner
Converts an idea into a structured development plan with milestones, tasks, dependencies, and implementation phases. Use when an idea or feature needs to be broken into a plan before work starts.
tech-debt-assessor
Inventories technical debt across a codebase and prioritizes it by cost-of-delay versus effort, producing a ranked paydown list. Use when debt is accumulating but it's unclear what to fix first.
api-designer
Helps design REST APIs - routes, request/response structures, validation, status codes, and conventions. Use when designing new endpoints or reviewing an existing API's design for consistency.