认证鉴权
认证与授权
browser
agentproto
Drive the user's REAL authenticated browser (their logins/cookies) via the local-browser chain — guilde → tunnel → daemon → chrome-devtools-mcp → cloned Chrome. The foundation skill for any "do X in a browser as me" task: navigate, capture (element screenshots + selective content/ld+json/readability), inspect network, manage tabs/cookies/consent, and — for data — talk to a site's OWN internal API instead of its fragile DOM. Site-specific recipes are separate skills (linkedin, x-twitter, youtube, instagram, tiktok, leboncoin, news-media) that build on this one.
ap-models-auth
agentproto
Manage models, billing profiles, and spend on the agentproto daemon — catalog_models for runnable routes, auth_profile_create/import/set_models for named credentials, harness presets, usage_rollup for spend estimates. Trigger when asked about models, costs, billing, auth profiles, OpenRouter routing, or how much a session spent.
x-twitter
agentproto
Drive the user's authenticated X/Twitter via the browser skill: read timelines/profiles, who-liked/retweeted (Favoriters/Retweeters), search, and gated posts/likes/DMs through X's internal GraphQL API. Use for any 'do X on Twitter/X as me' or build-the-X-graph task.
things-mac
nweii
"Manage Things 3 on macOS via CLI. Read, search, add, or update tasks, projects, and areas. Use for task management, listing inbox/today, and inspecting tags or repeating templates."
ap-tasks
agentproto
Use the agentproto shared task board — task_create, task_claim (CAS with rev), task_list, and task_update to coordinate parallel agents on one board, gate done with verify commands, and hand off work claimably. Trigger when coordinating parallel agents, splitting work across sessions, handing off tasks, or needing verifiable completion.
ap-spawn-agent
agentproto
Spawn a new agentproto agent session (claude-code, hermes, ...) with a model, cwd, prompt, and optional MCP tool access. Use when the user says "spawn an agent", "start a session", "launch claude-code on this task", "run hermes in the background", or "kick off a worker". Also covers listing already-running sessions before spawning a duplicate.
agentproto
Drive the user's real authenticated LinkedIn — via camofox stealth (preferred, distinct session + WhatsApp re-auth alert) or the chrome-devtools chain (fallback). API-first: read feed/profiles/activity, who-reacted/commented, connections, people + content search (who posted about a topic, with resolved permalinks), and build persona dossiers + engagement graphs — all FROM SOURCE via LinkedIn's Voyager API (page-context fetch), not fragile DOM. Plus gated write actions: message + attachment (API-first), inbox read, and the relationship graph (invite ±note / withdraw, via the trusted-input UI ladder since connect/withdraw are SDUI). Use for any "do X on LinkedIn as me" / "find + research + reach out to people" task.
supervise-long-missions
agentproto
"Route missions that outlive your attention — durable waiting, completion gates with retries, scheduled cron check-ins, shared task boards, spend tracking, and cleanup. Use when an agent works longer than one turn or you must not block on it. Triggers: supervise, long-running mission, check-in, non-blocking wait, completion gate."
ap-adapters
agentproto
Install and inspect agent CLI adapters (harnesses) on the agentproto daemon — adapter_list, adapter_install, harness_capabilities for credential/billing/tooling ground truth, role_list for spawn roles, plus the agentproto adapters CLI. Trigger when asked which agents are installed, to install claude-code or hermes, or to debug adapter credentials.
gathering-requirements
axiomantic
"Use when eliciting or clarifying feature requirements, defining scope, identifying constraints, or capturing user needs. Triggers: 'what are the requirements', 'define the requirements', 'scope this feature', 'user stories', 'acceptance criteria', 'what should this do', 'what problem are we solving', 'what are the constraints'. Also invoked by implementing-features during DISCOVER stage and by the Forged workflow."
agentproto-llm-endpoint
agentproto
Use the local LLM endpoint proxy (localhost:18090) — or your own gated public
ap-read-output
agentproto
Read what an agentproto session did — tail its recent output, export the full transcript, get a one-sentence summary, check cost/tokens, or locate a specific message. Use when the user says "what did the agent do", "show me the last output", "export that session", "how much did that session cost", "summarize what happened", or "did it finish or is it stuck".
archive-conversation
nweii
"Create analytical archival summaries of AI conversations, capturing intellectual journeys, key insights, and technical logs. Use when archiving, saving, or documenting a chat session."
update
rube-de
Audit open GitHub issues for staleness, orphaned blockers, missing labels, codebase drift, and already-implemented features. Categorizes issues into close-candidates, update-needed, needs-triage, stale, and healthy. Supports bulk and individual remediation via interactive approval. Triggers: audit issues, clean up issues, update stale issues, issue hygiene, backlog cleanup, prune issues.
github
Russ534
"GitHub via gh CLI: PRs, issues, reviews, repos, auth."
reinforce-skills
b-open-io
"This skill should be used when the user asks to 'reinforce skills', 'add skill map', 'update skill map', 'sync skills to CLAUDE.md', 'persist skills', 'save skills to project', 'embed skills', 'skills keep getting forgotten', 'I keep forgetting skills', or when setting up a new project where installed skills should be persisted as context. Generates a compressed skill-mapping directive in CLAUDE.md following the Vercel AGENTS.md research pattern."
bitable
yjwong
Access Lark Bitable databases - list tables, view fields, and read records. Use when user asks about a Bitable, database, or wants to query structured data.
notebooklm
b-open-io
This skill should be used when the user wants to query Google NotebookLM notebooks directly from Claude Code for source-grounded, citation-backed answers from Gemini. Provides browser automation, library management, and persistent auth. Drastically reduced hallucinations through document-only responses.
calendar
yjwong
Manage Lark calendar - view schedule, create/update/delete events, check availability, find meeting slots, RSVP to invitations. Use when user asks about meetings, schedule, availability, or calendar.
rulebook_task_update
hivellm
CLI tool to standardize AI-generated projects with templates, rules enforcement, and automation
create-next-project
b-open-io
This skill should be used when the user asks to "create a new project", "scaffold a Next.js app", "initialize a new app", "start a new project", "set up a new Next.js project", or mentions "create-next-project". Provides a guided, opinionated full-stack Next.js project initialization with Biome, Tailwind v4, shadcn/ui, better-auth, and Vercel deployment. Uses agent teams for parallel execution.
rulebook_task_create
hivellm
CLI tool to standardize AI-generated projects with templates, rules enforcement, and automation
contacts
yjwong
Look up employee information via Lark - find colleagues by ID, list department members, search users by name, search departments. Use when user asks about a person, colleague, job title, department, or org structure.
confess
b-open-io
"Systematic self-audit to discover and reveal mistakes, incomplete work, pattern violations, and hidden knowledge. Use when asking agent to 'confess', 'what did I miss', 'audit my work', 'what's wrong with this', or after completing significant changes."