Show HN: OJCP – an open protocol for agent-consumable job data

Hacker News Top Tools

Summary

OJCP (Open Job Context Protocol) is a new open standard built on MCP that lets AI agents discover, evaluate, and apply to jobs in a structured, consent-first way. It defines schemas and MCP tools for job search, application, and status tracking, aiming to standardize agent-job interactions.

Author here!<p>Agents are applying to jobs for people right now, with progressively more volume, and there&#x27;s nothing built for it. So they scrape career pages and fight ATS forms with Playwright&#x2F;Browser Use, which breaks constantly (or they get bot blocked). Employers get buried in applications that don&#x27;t fit, candidates hear nothing back, and the resume is now an AI-written thing that another AI scores (which breaks the existing model entirely, btw).<p>OJCP is MCP tools for search and apply, a manifest at &#x2F;.well-known&#x2F;ojcp.json so agents can find providers, and schemas that extend schema.org instead of replacing it. The playground on the site is a live MCP endpoint, so you can throw calls at it right now.<p>Why a spec at all when models keep getting better at figuring things out? Inference can&#x27;t produce authorization. An agent can work out what a form wants. It can&#x27;t establish that someone consented to this specific submission, and then employer has no way to verify who&#x27;s calling. So TL;DR a more capable agent is also a more capable impersonator.<p>In this model, trust runs both direction. Agents sign requests using the same method that CloudFlare and OpenAI are already using, providers sign their manifests, agents can check against a JWKS, and trust tiers cap how much candidate PII can go to a given provider. Validation happens at consent, so browsing costs nothing and you only pay the verify when the interaction occurs.<p>I&#x27;m the CTO of Recruitics (job advertising) and spent time at LinkedIn before that, so I&#x27;ve been at the intersection of hiring and job search for a while and have felt the pain of both sides.<p>Happy to answer any questions!
Original Article
View Cached Full Text

Cached at: 08/13/26, 06:21 PM

# OJCP — Open Job Context Protocol Source: [https://ojcp.dev/](https://ojcp.dev/) ![OJCP — Open Job Context Protocol](https://ojcp.dev/logos/primary-dark.svg) ## The open standard for \| OJCP defines how AI agents discover, reason over, and act on job opportunities\. Built on MCP\. Interoperable with schema\.org\. Governed by an[independent steering committee](https://ojcp.dev/governance), and designed to compose with the broader agentic web\. Why OJCP ## Built for the agentic web A shared protocol so AI agents and job providers speak the same language — structured, discoverable, privacy\-respecting\. MCP\-Native Tools OJCP tools are valid MCP tools\. Any MCP client can call search\_jobs, get\_job\_detail, and begin\_application out of the box\. Structured Discovery Providers expose a manifest at /\.well\-known/ojcp\.json\. Agents discover capabilities, tools, and apply paths automatically\. Normalized Apply Paths A standard taxonomy of application mechanisms — from ATS\-direct to agent\-ready flows — so agents know exactly how to apply\. Consent\-First Privacy Candidate data is opt\-in and scoped\. Resume embeddings enable fit\-scoring without transmitting PII\. Agents self\-declare identity\. Agent\-Aware Ranking Pass CandidateContext for personalized results\. Providers return fit\_score and fit\_rationale alongside standard job data\. WebMCP Ready Browser\-native agents can access OJCP tools via navigator\.modelContext\. Same schemas, same tools, new surface\. Specification ## Schemas & Tools OJCP defines seven core schemas and six standard MCP tools\. Providers MUST implement`search\_jobs`; all others are recommended\. Custom tools use namespaced names \(e\.g\.,`acme:get\_referral\_link`\)\. Discovery endpoint at /\.well\-known/ojcp\.json - `ojcp\_version` - `provider` - `tools` - `mcp\_endpoint` - `auth` - `rate\_limits` Extends schema\.org/JobPosting with agent\-specific fields - `ojcp\_id` - `skills\_required` - `experienceLevel` - `apply\_paths` - `urgency` Consent\-scoped candidate profile for personalized results - `consent\_scope` - `skills` - `experience\_years` - `employment\_type\_preference` Agent self\-identification for audit trails - `agent\_id` - `acting\_on\_behalf\_of` - `interaction\_mode` - `user\_consent\_token` Identity verification action for human completion - `step\_id` - `type` - `verifier\_id` - `verification\_url` - `human\_required` JWS proof with required claims: iss, aud, sub, iat, exp, nonce — no PII - `step\_id` - `verifier\_id` - `proof\_token \(JWS\)` - `subject\_hash \(SHA\-256\)` - `issued\_at` - `expires\_at` Discovery document hosted at /\.well\-known/ojcp\-verifier\.json - `verifier\_id` - `verification\_types` - `proof\_delivery\_methods` - `proof\_format \(jws\|jws\+jwe\)` - `signing\_algorithms` - `public\_keys\_url \(JWKS\)` How It Works ## Four steps from discovery to application Discover Agent probes`/\.well\-known/ojcp\.json`or queries the OJCP Registry to find providers\. The manifest declares available tools, apply paths, and authentication requirements\. Search & Evaluate Agent calls`search\_jobs`with optional`candidate\_context`for personalized ranking\. Each result includes`fit\_score`, apply paths, and structured metadata\. Apply Agent calls`begin\_application`with an`agent\_declaration`and the candidate's consent token\. The provider returns required fields and a session for multi\-step flows\. Track Agent calls`check\_application\_status`to monitor progress\. Providers emit structured status updates throughout the hiring pipeline\. Examples ## See it in action From discovery to application — real OJCP payloads your agents can use today\. Apply Paths ## Normalized application taxonomy OJCP standardizes the fragmented landscape of application mechanisms into a taxonomy agents can reason over\. TypeDescriptionAgent Submission`ats\_direct`Apply directly via ATS \(Workday, Greenhouse, Lever, etc\.\)Varies `provider\_hosted`Provider controls the apply flow and delivers to ATSFull Support `platform\_native`Third\-party platform owns the flow \(e\.g\., Indeed Apply, Easy Apply\)Limited `email`Legacy email\-based applicationNot Supported `external\_redirect`Redirect to opaque external pageNot Supported `custom`Catch\-all for non\-standard apply mechanisms \(homegrown, conversational, etc\.\)Varies Live Playground ## Try the OJCP API This site is a live OJCP provider\. Fire real tool calls against mock data — the same endpoint MCP clients connect to\. MCP: /api/mcp Manifest: /\.well\-known/ojcp\.json

Similar Articles

Ask HN: Who is using MCP in production?

Hacker News Top

A Hacker News user asks about production usage of Model Context Protocol (MCP), with comments discussing its advantages and examples from real-world applications like UI updates and open data initiatives.

Open Computer Use

Product Hunt

Open Computer Use is an open-source MCP (Model Context Protocol) for AI agents to control computer interfaces.