Vendor: Cyan Eclipse Technologies · Platforms: MCP over stdio
Certification scope
Test Bench v0.1.0, 2026-10-25, SANDBOX environment: policy refusals (blocklist, salary floor, daily cap, expired policy), pause/kill switch, zero-fact tailoring diff, fabrication catch, hash-chained receipt ledger, dry-run transmits nothing, delete_my_data purge — 17 pass, 1 na of 18 checks
What it does
- Apply to jobs only inside a human-granted written policy: target roles, salary floor, locations, company allowlist + blocklist, daily cap, approval threshold, expiry, kill switch
- Live job discovery against the public Greenhouse / Lever / Ashby / Workable JSON feeds
- Review-first by design: matches land as prepared applications for human approval
- Tailored resumes re-order verbatim real experience only — every tailoring ships with a machine-verifiable diff proving zero invented facts
- Append-only, hash-chained receipt ledger for every grant, preparation, approval, refusal and submission
- v0.1 submissions are DRY-RUN ONLY: the Greenhouse payload is built, validated and recorded — nothing is transmitted to any employer
Tools
ingest_resumeset_policyget_policysearch_jobsmatch_jobsprepare_applicationapprove_applicationreject_applicationlist_receiptsget_trackerpauseresumedelete_my_data
Problems it addresses
- I want help applying to jobs, but only to roles I actually want, at or above my salary floor
- Auto-apply tools invent experience on resumes — I need proof mine never does
- Anything applying in my name needs a blocklist, a daily cap, and a kill switch
- I want a receipt for every application prepared, approved, refused, or submitted in my name
Not tested
- Auth enforcement (bench U1 = na: local stdio transport has no credential layer to test)
- Live job boards — bench jobs are synthetic fixtures seeded into bench state in /tmp; no live boards fetched, no employer contacted
- Live submission — v0.1 is DRY-RUN ONLY by design; A7 truthfulness check ran logic-level (verify_truthful() in-process), all other checks over MCP
Bench scorecard: test-bench/results/eclipse-apply-mcp-2026-10-25.json
Evaluation-intake listing
This listing records the tool's place in the Bazaar's evaluation
and monitoring pipeline, with the evidence gathered so far printed above — verdict, scope,
and not-tested items included. It is not sold on this site, and its presence in the catalog
is not a claim beyond that evidence. The service behind every verdict is the bench:
what a bench run buys →