Skip to content

Latest commit

 

History

History
77 lines (49 loc) · 4.37 KB

File metadata and controls

77 lines (49 loc) · 4.37 KB

noticed-skills

A suite of agent skills for noticed — a personal networking agent. Each skill wraps the noticed MCP to handle one job in the relationship-management loop: capturing people you meet, debriefing meetings, researching contacts, following up, searching your network, and onboarding.

connect noticed first

These skills call the noticed MCP. Connect it to your client before you install the skills, or they will not have anything to talk to.

Go to noticed.so/onboarding and pick your client — Claude, Claude Code, Cursor, Gemini CLI / Antigravity, ChatGPT, or anything else that speaks MCP. The MCP server itself lives at noticedso/cli if you want to run it locally.

install (Claude Code)

/plugin marketplace add noticedso/skills
/plugin install noticed-skills@noticed-skills

install (claude.ai)

Requires a Pro, Max, Team, or Enterprise plan with code execution enabled (Settings → Capabilities).

  1. Open Customize → Plugins and click + → Add marketplace.
  2. Sync from this GitHub repo: noticedso/skills.
  3. From the synced marketplace, install noticed-skills. All six skills land in one shot.

install (ChatGPT)

ChatGPT has no GitHub marketplace install, so each skill goes in by hand.

  1. Open Skills → New skill → Upload from your computer.
  2. Upload each skills/<name>/ folder from this repo as its own skill (zip the folder first if ChatGPT asks for a single file). Repeat for all six.

Workspace admins: confirm Enable skills and Enable skill uploading are on under Permissions & roles.

the skills

skill what it does writes?
remember-person Capture one or more people into noticed from a name, URL, handle, photo, or freeform dump. The core capture primitive. yes
event-debrief Process a meeting/event dump into people, notes, follow-ups, ideas, and references — all written in one pass. yes
research-person Deep-dive one person before a meeting; render a dossier from noticed + the web; offer to save it back. only on save
follow-up Draft a follow-up message to someone you just met, in your voice, with something actionable in it. logs the touchpoint only after you confirm it went out
search-network Natural-language search across your network; returns a tight table; drills into anyone. read-only
interview-me Onboarding questionnaire that captures your identity, goals, and current tools. yes (to your own record)

how they fit together

The capture loop runs across a real-world interaction:

remember-person  →  event-debrief
   (during)            (after)

research-person and follow-up support any contact at any time. search-network is the read path over everything you've captured. interview-me is first-run setup.

shared conventions

All skills follow the same rules, so they read as one system:

  • Identity resolution is owned by remember-person (own-network search → web enrichment → ask). The other skills follow that flow rather than re-implementing it.
  • Provenance, two surfaces. In chat, web-found facts are attributed softly ("his linkedin says…") so research never reads as something you said. In the stored note, every line carries a system-only tag ([from user] / [research, unverified]) that is never shown in chat.
  • Readback after every write. noticed writes are silent, so each skill recaps what landed — in prose, not a database-style tags: row.
  • No public scope in the capture path. Capture answers who is this person via own-network + web. Warm-intro / reachability discovery (scope: "public") lives only in search-network, where it falls back to own-scope silently if the backend errors.
  • No Gmail dependency. Drafts are copy/paste (subject included for email); nothing is sent automatically.

testing

See TESTING.md for suggested prompts to exercise each skill and a checklist of behaviors to confirm.

updates

Updates are manual for now. Reinstall to get the latest.

notes

Design notes and per-skill detail live in the Notion workspace under 🤹 mcp: skill ideas. These skills were revised and tightened during NY Tech Week 2026 against the live MCP learnings; product gaps that surface during use go to the 🔧 NYTW: mcp learnings page.