メニュー
The security review checklist (risk / validation / abuse / data) with this repo's anchors. Load when reviewing security-sensitive changes — capabilities, IPC, deps, secrets, AI, privacy, rate-limits.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
Shared write-side contract every domain AUTHOR imports — how to implement reliably, the smallest-diff rule, validation gates, and the never-approve-your-own-work rule. The write-side mirror of token-efficiency. Load at the start of any implementation task.
Shared context-discipline contract every agent imports — context-source priority, read budget, confidence-stop, the severity rubric, and terse output. Load at the start of any review or implementation task.
General repo rules enforced by ESLint/TypeScript/CI — PRs-only, never bypass lint, import ordering, type imports, path privacy, commit format. Load for any code change.
Browser-extension + bridge standards — MV3 (Chrome + Firefox), least-privilege permissions, the loopback pairing-token auth model, the shared TS↔Rust wire protocol, and the Chrome Web Store + Firefox AMO store-policy + pre-submission checklist. Load for changes under apps/extension/**, extension_bridge/, and the extension protocol.
React renderer standards — ports & adapters (service hooks), design tokens, motion tokens, @ajh/ui primitives, i18n, feature isolation, React Query, a11y. Load for changes under apps/desktop/src/renderer/**.
Rust/Tauri backend standards — L0–L3 layers, centralized platform/net/error ownership, Rust-first business logic, registries, error handling. Load for changes under apps/desktop/src-tauri/src/**.
| name | security-checklist |
| description | The security review checklist (risk / validation / abuse / data) with this repo's anchors. Load when reviewing security-sensitive changes — capabilities, IPC, deps, secrets, AI, privacy, rate-limits. |
Authoritative: docs/knowledge/security-rules.md. Severity bias for security/data findings = round UP.
tauri.conf.json CSP intact (incl. Ollama 127.0.0.1:11434) · capabilities/default.json least-privilege · updater signing key + latest.json integrity · deny.toml / cargo audit / pnpm audit clean · new deps license-checked.Re-baseline to current editions: OWASP Top 10 2025 (was 2021), ASVS 5.0 (was 4.x), SLSA v1.2 (was v1.0), LLM Top 10 2025, CWE Top 25 2025. Cite these, not older ones.
A03 Software Supply-Chain Failures (expands old A06) and A10 Mishandling of Exceptional Conditions; SSRF folded into A01; A02 Security Misconfiguration rose to #2.pnpm-lock.yaml + CI --frozen-lockfile; cargo audit/cargo deny vs RUSTSEC daily. ⚠️ Provenance ≠ safety — the 2025–26 Shai-Hulud npm worm shipped malware with valid SLSA L3 provenance via hijacked OIDC tokens → pin exact versions, --ignore-scripts, scope-lock publish tokens. https://unit42.paloaltonetworks.com/npm-supply-chain-attack/tauri-plugin-stronghold/keyring), never config/logs. Tauri least-privilege capabilities + strict CSP + minisign-signed updater (any manifest/key compromise = full takeover). https://v2.tauri.app/security/Common mistakes: citing Top 10 2021 / ASVS 4.x / SLSA v1.0 (all superseded); treating npm/SLSA provenance as a safety guarantee (Shai-Hulud defeated valid L3); wildcard Tauri capabilities or no CSP ("it's a desktop app"); secrets in tauri.conf.json/env/logs; lockfile uncommitted or CI without --frozen-lockfile; no cargo audit/cargo deny gate; ignoring A03/A10 + LLM01/LLM06 for AI features.