Where your career already lives.One context layer for every agent.
ProfileClaw serves your consented career context straight into the agents you already run — over REST or MCP, scoped to exactly what you allow.
The agent directory
Every client below reads the profile layer over API or MCP — each scoped to exactly the layers the user consented to share.
Coding agents
Pair programmers that read skills and work style.
Claude Code
Codex
Cursor
GitHub Copilot
ZCodeAssistants
General assistants grounded in verified career context.
ChatGPT
Claude
Gemini
PiAutonomous agents
Long-running agents acting inside profile guardrails.
Devin
Manus
OpenClaw
WorkBuddyResearch & harnesses
Answer engines and model harnesses reading scored dimensions.
Perplexity
DeepSeek
HermesA context API agents can reason over
Read raw answers, dimension scores, or narrative summaries as separate layers — scoped to exactly what the user consented to share with your client.
GET /v1/context?layers=scores,summaries
Authorization: Bearer <client-token>
→ 200 OK · scope: scores,summaries
→ communication 0.82 · writing 0.74 · leadership 0.68
Plug in over MCP
ProfileClaw ships an MCP server, so agent clients can read career context as native tools — no custom integration code.
Connect the server
Add ProfileClaw to your client's MCP config and sign in once.
Grant a scope
The user picks which layers your client may read. Everything else stays sealed.
Read context
Call the tools when you need them — fresh, consented, and logged.
Claude CodeCursorCodexOpenClawWorkBuddyIntegrate once —
every agent starts reading.
Create an API key
Issued in your dashboard, scoped per use
Wire REST or MCP
One endpoint, two protocols
Agents start reading
Claude Code, Cursor, OpenClaw…