ProfileClaw
15 min · build once · agents read forever

Your AI Already Knows What to DoProfile is the new protocol

It just needs your profile. Build it once, let every agent start in context.

Profile first, action after4 assessment tracks · 200+ questionsWorks with OpenClaw, Claude Code, and next-gen agents

Encrypted by default · Your key stays yours · You choose which agents can read

Profile_Protocol_v2

Your profile

Human signal layer

Compounding

Assessment

Interests and work style

Resume

Experience and signals

Boundaries

Values and constraints

Profile loaded

Action starts in context

Ready for agents

Strengths, preferences, and limits loaded.

1
High-fit opportunities
2
Context-aware writing
3
Action with memory
Mem_State

Profile ready

Context that persists

Exec_Engine

In motion

Context becomes action

Connect once. Let agents keep going.

Paste one instruction. It loads your profile and suggests the next best moves.

Paste to your agentRecommended
Agent Prompt
>Read https://profileclaw.com/en/skill.md, connect my ProfileClaw, then continue helping me from my full profile.

Best for OpenClaw, Claude Code, OpenCode, and other runtimes that can continue automatically.

CLI fallback
Terminal
$npx @profileclawai/skill install-skill

Use this if your runtime cannot read the skill file directly.

Turn your past into signalMake experience readable by agents

Turn scattered experience into structured context. Then agents act from the same version of you.

Let AI read real experience

Projects, skills, and industry background become structured data, so AI reads you faster.

Multi-turn chat burns too many tokensThat is why profiles need compression

ProfileClaw gives agents stable context first, so tokens go into execution instead of rediscovery.

Explain yourself less

Individual users

A reusable source of long-term context

Frequent agent users

A human-context layer that plugs into internal agent workflows

Teams and enterprises

Multi-turn chat burns too many tokens
Build once, read everywhere

Build the foundation in 15 minutesSee yourself clearly, then let agents build

Once your assessments and materials are in place, you get a continuously updated profile for judgment and workflows.

Localized assessments
Four assessments cover interests, personality, values, and skill signals for real work in China.
Profile
Assessment results, resume evidence, and materials become one structured profile that you and AI can read.
AI analysis
Your real signals become direction and next steps instead of generic advice.
Agent workflows
With OpenClaw, Claude Work, or Nanobot connected, AI can move work forward without making you repeat yourself.
Developer access
Generate a ProfileClaw Key so approved tools can read the profile in a controlled way.
Broader opportunity exchange
The same profile supports jobs, freelance work, collaborations, creative paths, and long-term exploration.

Profile layer

High-quality input creates high-efficiency outputLet token efficiency compound

ProfileClaw turns assessments, materials, and preferences into a structured profile. You read the map. Agents read context.

Clarify yourself first

Turn your interests, experience, and strengths into a structured context asset.

25%

步骤 1

Keep the context evolving

Add new experience, skills, or material. The context updates with you.

0%

步骤 2

Use tokens where they matter

The next time you call AI, tokens go into moving work forward instead of re-explaining who you are.

0%

步骤 3

Let context drive real action

Job search, freelance work, collaboration, or long-term exploration—agents keep acting from the same context.

0%

步骤 4

Use cases

Start with job search, grow beyond itOne profile, broader decisions

Job search is the clearest starting point. Once your history, strengths, and limits are structured, the same profile supports broader moves.

Job search and transitions

Give agents a stable profile for role search, fit judgment, and transitions.

Role filteringTransition judgmentInterview positioning
Collaborations and projects

Agents can judge people, projects, pace, and boundaries so you screen fit earlier.

Collaboration fitRole splitBoundary clarity
Freelance and client work

Pass your skills, boundaries, and preferences to an agent so it can screen projects and draft proposals.

Project screeningProposal draftingPricing support
Content and personal brand

Use your lived experience and durable themes to give agents a stronger base for content direction.

Content positioningTopic planningKnowledge products
Advisory and services

If you already monetize judgment or delivery, this layer helps agents package offers and filter clients.

Offer packagingSolution draftsClient matching
Recruiting and talent ops

For organizations, the same context layer supports recruiting, internal mobility, talent review, and agent workflows.

Hiring filtersInternal mobilityTalent review

Start with the most immediate use case. As the profile grows, it supports more decisions.

Capture oncethen connect every next step

Once the profile is ready, ProfileClaw keeps your history, preferences, and signals in a structure agents can reuse.

Get your API key

Generate a key in settings, then start with the Context API.

01

Individuals and builders

When your profile is connected to OpenClaw, Claude Work, Nanobot, or your own workflow, agents receive structured experience, preferences, limits, and signals.

02

Enterprise recruiting (in progress)

At scale, employer workflows use the same context layer, with user permission, for matching and talent ops.

Security and access control

Every call is authenticated and encrypted. Access stays user-controlled and revocable.

Sync the profile with an agent
Let AI handle first-pass outreach
Explore directions from your boundaries
Keep guidance and next actions updated
Capture once

Once the profile is ready, ProfileClaw keeps your history, preferences, and signals in a structure agents can reuse.

then connect every next step

Once the profile is ready, ProfileClaw keeps your history, preferences, and signals in a structure agents can reuse.

Assessment completed
Assessment completedJust now
Resume uploaded
Resume uploaded2 min ago
Profile created
Profile created5 min ago
Agent workflow connected
Agent workflow connected8 min ago
Assessment completed
Assessment completedJust now
Resume uploaded
Resume uploaded2 min ago
Profile created
Profile created5 min ago
Agent workflow connected
Agent workflow connected8 min ago
Assessment completed
Assessment completedJust now
Resume uploaded
Resume uploaded2 min ago
Profile created
Profile created5 min ago
Agent workflow connected
Agent workflow connected8 min ago
High-priority opportunities filtered5 directions
Personal pitch generated3 drafts
Potential collaborators identified2 profiles
Next action updatedThis week
High-priority opportunities filtered5 directions
Personal pitch generated3 drafts
Potential collaborators identified2 profiles
Next action updatedThis week
High-priority opportunities filtered5 directions
Personal pitch generated3 drafts
Potential collaborators identified2 profiles
Next action updatedThis week
职虾

First, they made themselves legibleThen better opportunities became visible

These users used ProfileClaw to clarify direction before making decisions.

I stopped applying harder. I started seeing clearer. Once my fit became legible, the right interviews appeared.
Xiaolin · Recent graduate

Xiaolin · Recent graduate

Now: Market Analyst · RMB 12K/month

The profile showed me I was not stuck. I was misaligned. Once that became obvious, the next move was easy.
Zhang · Career switcher

Zhang · Career switcher

Now: Technical Pre-Sales · 400K+ RMB/year

My background was real. It was unreadable. Once the profile translated it into local signal, the market read me correctly.
Jessica · Returnee

Jessica · Returnee

Now: HRBP · 350K RMB/year

Frequently asked questions

If this is your first time seeing ProfileClaw, start here.

ProfileClaw FAQ

What it is, who it is for, and how to begin.

ProfileClaw is a long-term profile built by 职虾. It stores assessment signals, experience evidence, skill patterns, and preferences so you see a readable map while AI sees stable context.

15 minutes · one-time setup · long-term reuse

Profile first, then actionBuild once, let value compound

Use 15 minutes to build your profile. After that, agents can spend tokens on real work instead of starting from scratch.

Build once, let value compound