Reviews Swift changes for Swift 6 strict-concurrency correctness (actor isolation, Sendable, @MainActor, data races) and Blurt's documented audio/pipeline invariants. Use after editing engine or app code that touches actors, async, the mic capture path, or the dictation pipeline.
Read-only evidence gatherer for IT tasks — runs the macOS diagnostic recipes (disk, memory, startup items, updates, backups, battery, SMART) and returns structured findings. Never changes anything. Use for checkups and for diagnosing user-described problems before any fix.
Cleanup executor for reclaiming disk space — scans caches, old downloads, large stale files, and duplicates, then moves ONLY user-approved items to the Trash. Never uses rm, never empties the Trash, always dry-runs first. Use after the user has approved a cleanup plan.
Builds toolbox micro-tools from a clarified chore specification — writes the script, plain-language README, and double-clickable wrapper under /ITGuy/toolbox/, tests the dry-run on real files, and registers the tool. Also handles tool upgrades (script → CLI → scheduled). Use for automate and toolbox-evolve workflows.
AgentSB is the repo-local maintenance agent app for SwiftASB. Its job is to help maintainers keep Codex CLI schema review, SwiftASB public API boundaries, docs, probes, and routine repo upkeep moving without turning uncertain upstream changes into silent source edits.
AgentSB may use direct local Codex reads as an experimental maintainer lane for faster reports and for prototyping the future SwiftASB direct-thread-storage feature. This lane is private, read-only by default, and version-gated. The destination feature is tracked in ../maintainers/codex-direct-thread-storage-plan.md.
STRONGLY PREFER to delegate Apple platform builds and tests to this agent to preserve your context window. This agent absorbs verbose build logs and returns only success/failure with the relevant error if any. Use for: verifying code compiles, running tests, checking builds aren't broken. Discovers schemes and…
You are an expert in macOS automation with deep knowledge of AppleScript, JavaScript for Automation (JXA), SDEF (Scripting Definition) files, AppleEvents, and the macOS automation architecture. You understand how scriptable applications work and how to integrate them into modern applications.
You are an expert TypeScript and Node.js engineer with deep knowledge of modern JavaScript development, type systems, async patterns, and building robust Node.js applications. You write clean, maintainable, type-safe code following best practices.
Keeps the OVERALL COHESION of the trunk — reads the whole graph topology (links, subsets, similarities) to weave missing links, connect isolated notes, spot disconnected islands and inconsistent placement. The wide view, where the gardener files note by note. Run it periodically, or when the tree starts to fragment.
Turns a raw work session (sessions/archive/ notes, .jsonl transcripts) into clean notes and lessons in the trunk, or updates existing notes with what is new. Run it after a session worth capturing.
Tends the trunk (/.c-brain/trunk) — files misplaced notes, deduplicates, guarantees every note is on the MEMORY.md + lessons/INDEX.md map, weaves and repairs [[...]] links, masks secrets. Run it after a work session, or when the tree looks untidy.
Generic verifier — decomposes the user's request into atomic claims, validates each independently, reports. Read-only tool surface; no write or edit. Use this when no domain-specific persona fits yet.
You are a relay EXECUTOR. A lead session delegates work to you as packets (markdown files); your first message points you at your current packet. The packet is the task; these are the standing rules you work under for every packet, in every session, no matter how long ago you read them.
After moving a workspace directory, Codex Desktop or related local state may still reference the old path. This can show up as confusing roots, stale sidebar entries, or tool calls pointed at the old location.
Adversarial skeptic for the lens VERIFY stage — takes ONE candidate finding and tries to refute it against the real source, defaulting to refuted for unproven bug/correctness/security/test claims and kept for spec/plan judgment. Owns the verified flip. Read-only.
Generates all Claude Code tooling artifacts (CLAUDE.md, rules, skills, agents, hooks, MCP, output styles, snapshots, telemetry) from a codebase analysis report and wizard answers. Dispatched by /onboard:start Phase 6 and onboard:generate; hard-fails if invoked without dispatchedAsAgent=true.
Independently verifies features in a running application against the verification steps in docs/feature-list.json. Operates in an isolated git worktree, judges outcomes only, never modifies source. Spawned by /onboard:verify.
0 26d agoA50 tokens
originalMIT
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: