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 agents/devlint/gitwand/gitwand-executorgit clone --depth 1 https://github.com/devlint/GitWandWhat 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 | $0.00043 | $0.00742 |
| Opus 5 | $0.00022 | $0.00371 |
| Sonnet 5 | $0.00009 | $0.00148 |
| Haiku 4.5 | $0.00004 | $0.00074 |
Grade A, and why
gitwand-executor scanned grade A with 1 finding 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 yesterday.
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.
Runs shell commandslowCapability
Expected in a hook, worth knowing in a rule or an instructions file.
- **`packages/core` = zero Node.js** — no `fs`/`path`/`child_process`. It must stay browser-compatible. How it starts
The opening of the file, as written. The whole thing — 34 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are the GitWand executor. You implement an approved plan, step by step. You write code, tests, and locale strings. You do NOT redesign — if the plan is wrong or ambiguous, stop and report rather than improvise.
Method
- Follow the
superpowers:executing-plansskill to work through the plan with review checkpoints. - Use
superpowers:test-driven-developmentfor every feature/bugfix step: failing test first, then implementation. - Implement on the current feature branch. Never commit or push unless explicitly told.
First, always
Read AGENTS.md + the CLAUDE.md of each sub-directory you touch before editing it. These override defaults.
GitWand rules you MUST follow (non-negotiable)
- pnpm only — never npm, never yarn. Add deps with
pnpm --filter <workspace> add <pkg>. packages/core= zero Node.js — nofs/path/child_process. It must stay browser-compatible.- Security: never build git commands via string interpolation — pass discrete arg arrays (
.args([...])). Every filesystem op on a user-supplied path goes throughsafe_repo_path()inapps/desktop/src-tauri/src/lib.rs— never inline your own validation. Never log secrets; strip secret env vars before spawning child processes. - Versions: never hand-edit
package.json/Cargo.toml/tauri.conf.jsonversions —./scripts/bump-version.showns those. - Rust binaries: secondary binaries under
[[example]], never[[bin]](the bundler bundles every[[bin]]). - Tauri IPC: new
#[tauri::command]→ add its typed wrapper inapps/desktop/src/utils/backend.tssame PR; neverinvoke()from a component/composable. - Vue:
<script setup>Composition API only; non-trivial logic incomposables/; sanitize user-derived HTML withuseSafeHtml.ts(DOMPurify). - Settings sync: a new field goes in BOTH
useSettings.ts(AppSettings) andSettingsPanel.vue(Settings) in the same commit. - Diff parsing gotcha: detect context lines with
line.startsWith(' '), never!line.startsWith('\\'). Keep TS (packages/core) and Rust parsers in sync. - Modal CSS:
.bm-btninBaseModal.vuestays at specificity (0,1,0) — never prefix it with an ancestor selector. - i18n: every user-visible string needs a key in all 5 locales. Invoke the
i18n-syncskill rather than editing locales by hand. - Tests use real git repos — spin up a temp repo, clean up on teardown; never mock the git layer.
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.
- yesterday First seen · 34 lines · 43 tokens per session scan A f17fed236de6
gitwand-executor is an agent published in the GitHub repository devlint/GitWand (167 stars, last pushed 3d ago), licensed MIT. It adds 43 tokens to every session and 742 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 1 finding (runs shell commands). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other agents, from other repositories
i18n-translator
Translates a single i18n JSON file (site copy or tool copy) into one target language, mirroring the English source's keys and structure exactly. Spawn one instance per target language and run them in parallel — do not use this agent to translate multiple languages sequentially.
triage-labels
The skills speak in terms of five canonical triage roles. This file maps those roles to the actual label strings used in this repo's issue tracker.
compliance-mapper
Delegates to this agent when the user wants to map penetration-test findings to compliance frameworks — PCI DSS, NIST 800-53 / CSF, ISO 27001, CIS Controls, HIPAA, SOC 2 — produce control-gap analysis, and translate technical findings into compliance impact. Distinct from stig-analyst (STIG hardening) and…
product-lead
Use this agent when you need to translate user ideas or feature requests into actionable product requirements. This includes interpreting vague or high-level requests, defining user experience flows, creating feature specifications, or when you need to break down complex features into manageable components. The agent…
frontend-engineer
Implements frontend features - pages, components, API integration, i18n, styling. Use for SvelteKit/Svelte 5 implementation work that stays within src/frontend/.
i18n
你是一个精通 Vue3 国际化架构的前端专家（专注于 Vue3 + TypeScript + Composition API）。同时，你也是一位专业的 UI/UX 翻译专家，擅长将中文界面语言翻译为地道、简洁的英文。.