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

A plan to let AI job-hunt for you sparked instant eye-rolls, jokes, and rage

TLDR: OJCP is a new shared format meant to help AI assistants search and apply for jobs across different hiring sites. Commenters were less impressed by the tech than by the bigger problem: hiring is already cold and broken, and many want accountability for companies that ghost applicants.

A new project called OJCP wants to make job listings easier for AI assistants to read, compare, and even apply to on your behalf. In plain English: employers, job boards, and hiring software companies are trying to create a shared rulebook so bots can browse openings, check details, and start applications without getting lost in a maze of forms. Supporters see a cleaner, more organized future for online hiring. The comments? Much messier, much louder, and way more fun.

The biggest eye-roll came from people who think this is wildly premature. One commenter basically said, “Can we please stop calling every early draft a protocol?” That set the tone: less polite applause, more skepticism about tech people slapping grand names on unfinished ideas. Others went straight for the existential dread, joking that they first thought “agent-consumable job data” meant AI would do the jobs, not just apply for them. And then came the real anxiety bomb: if hiring already feels impersonal, what happens when the process becomes “AI said this guy” on both sides?

But the most emotionally charged reactions weren’t about standards at all — they were about the broken reality of job hunting. One of the sharpest suggestions was to forget fancy AI pipelines and instead build a public blacklist of companies that ghost applicants the most. That struck a nerve, because behind all the buzzwords was a brutally relatable complaint: people are sending out hundreds of applications and hearing nothing back. So while OJCP pitches a smoother future for machine-friendly hiring, the crowd’s verdict was clear: cool idea, but can someone fix the human misery first?

Key Points

  • OJCP is presented as an open protocol for AI agents and job providers to exchange structured job and application data.
  • The protocol uses MCP-compatible tools and a discovery manifest at `/.well-known/ojcp.json` so agents can find provider capabilities, tools, apply paths, and authentication requirements.
  • OJCP defines seven core schemas and six standard MCP tools, with `search_jobs` required for providers and other tools recommended.
  • The specification emphasizes consent-scoped candidate data, agent self-identification, resume embeddings for fit scoring without transmitting PII, and verification through JWS-based proofs.
  • The workflow described in the article covers discovery, job search and evaluation, application initiation, and status tracking, alongside a standardized taxonomy of application paths such as `ats_direct` and `provider_hosted`.

Hottest takes

"I find it offensive to use the word protocol" — sandeepkd
"I was hoping this was for agents to consume job data so they can perform the job" — fidotron
"track which companies reject/ghost the most applicants" — zackmorris
Made with <3 by @siedrix and @shesho from CDMX. Powered by Forge&Hive.