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 skills add photostructure/coding-skills --skill gitplangit clone --depth 1 https://github.com/photostructure/coding-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/photostructure/coding-skills/gitplan)<a href="https://agentmods.dev/skills/photostructure/coding-skills/gitplan"><img src="https://agentmods.dev/badge/skills/photostructure/coding-skills/gitplan/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/photostructure/coding-skills/gitplan"><img src="https://agentmods.dev/badge/skills/photostructure/coding-skills/gitplan.svg" alt="Reviewed on agentmods" width="80" 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 5.1 | $0.00033 | $0.01274 |
| Opus 5.5 | $0.00013 | $0.00510 |
| Sonnet 5.5 | $0.00007 | $0.00255 |
| Haiku 4.5 | $0.00003 | $0.00127 |
Grade A, and why
gitplan 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 14d 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 — 80 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Applicability is about how tangled the changes are, not how many files they touch. Invoke this skill when the working tree mixes multiple intertwined logical changes that need untangling into separate commits — even if that's only a handful of files. Skip it when the changes are trivial or superficial, no matter how many files they touch: formatter runs, lint autofixes, typo/grammar fixes, or edits that obviously belong in a single commit.
Review and plan git commits
More on these workflows: photostructure.com/coding
User direction
User instructions take precedence over this skill's grouping, review, and approval defaults. If the user says "commit" or "approved, commit" for an understood scope, use stage to verify and commit that scope. Unless the user also requests review first, do not restart planning, invoke review skills, wait for a second opinion, or ask for approval again. This rule also applies when the instruction arrives during the workflow below. When project instructions require review before every commit, a commit instruction for changes that review has not finished on since they last changed runs the review first, and only an explicit instruction to skip review skips it.
Propose focused, coherent, reviewable commits. Honor grouping the user has already specified or approved.
If the repository has a layered structure (e.g. shared utilities → core → feature packages → app), work through it from the lowest-level layer upward so dependencies are committed before their consumers.
Workflow
Phase 1: Identify Themes
- Scan all current changes with
git statusandgit diff --stat. When subagents are available and the diff is large, use them to preserve context and summarize distinct areas. - For complex diffs, use
git diff -U150but limit JSON/lockfiles to the first ~50 lines. - Identify logical themes/groupings. Each theme must have a single coherent purpose — a unifying "why" that explains every file in the group. If you can't state the purpose in one sentence without using "and", split the theme. Never create catch-all buckets like "housekeeping", "misc", "cleanup", or "various fixes". Every file belongs in a theme because of what it does, not because it's small or doesn't fit elsewhere. Orphan files that truly don't relate to any theme get their own single-file commit.
- Bundle related docs/plans with their code changes. If a planning doc, design note, or task file corresponds to a theme, commit it alongside the code it describes — never lump it into a separate "docs" commit. Docs that don't correspond to any code change can go in a docs-only commit.
- Present the themes to the user as a numbered list with brief descriptions. Order by increasing complexity/risk.
- Ask: "Which theme should we focus on first?"
What ships with it
1 file 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.
- 14d ago Changed · +6 lines a3af284f28da
- 28d ago Changed · -1 lines b5ddddf3f395
- 1mo ago Changed · +12 lines 2d7ea2af8bc3
- 1mo ago First seen · 63 lines · 33 tokens per session scan A 36b9c9279317
gitplan is a skill published in the GitHub repository photostructure/coding-skills (4 stars, last pushed 13d ago), licensed MIT. It adds 33 tokens to every session and 1,274 once invoked, about $0.0001 per session on Opus 5.5. 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
pr-review
Address feedback left on a GitHub pull request: fetch unresolved review threads, make agreed Elixir/Phoenix code fixes, reply, and resolve. Use for a PR URL/number or reviewer comments. NOT for pre-PR review, findings triage, or CI monitoring.
phx-pr-review
Address feedback left on a GitHub pull request: fetch unresolved review threads, make agreed Elixir/Phoenix code fixes, reply, and resolve. Use for a PR URL/number or reviewer comments. NOT for pre-PR review, findings triage, or CI monitoring.
signing-preflight
Before the first commit of a session, check that commit signing will work — the key is loaded and unlocked — and ask the operator to unlock it once, up front, instead of discovering it after a hundred unsigned commits. Never disables signing silently, never re-signs history without asking. Use at the start of…
github-pr-creation
Creates GitHub Pull Requests with automated validation and task tracking. Use when user wants to create PR, open pull request, submit for review, or check if ready for PR. Analyzes commits, validates task completion, generates Conventional Commits title and description, suggests labels. NOTE - for merging existing…
git-commit
Creates git commits following Conventional Commits format with type/scope/subject. Use when user wants to commit changes, create commit, save work, or stage and commit. Enforces project-specific conventions from CLAUDE.md.
phx-watch-pr
Watch an Elixir/Phoenix PR with an Amp Orb keep-alive lease until required non-deployment CI is green and review threads are resolved.