Borrowing it
Nothing to install: this file belongs to amikai/openings-mcp. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/amikai/openings-mcp/main/.agents/skills/discover-companies/SKILL.mdgit clone --depth 1 https://github.com/amikai/openings-mcpWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/amikai/openings-mcp/discover-companies)<a href="https://agentmods.dev/skills/amikai/openings-mcp/discover-companies"><img src="https://agentmods.dev/badge/skills/amikai/openings-mcp/discover-companies/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/amikai/openings-mcp/discover-companies"><img src="https://agentmods.dev/badge/skills/amikai/openings-mcp/discover-companies.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00082 | $0.01203 |
| Opus 5.5 | $0.00033 | $0.00481 |
| Sonnet 5.5 | $0.00016 | $0.00241 |
| Haiku 4.5 | $0.00008 | $0.00120 |
Grade A, and why
discover-companies scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 100 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Discover Companies
Overview
Given one roster-based provider, source companies hosted on that ATS
that no roster has yet, confirm each against the live adapter path, and
land them in internal/provider/<provider>/companies.yaml. This skill
covers the sourcing; verification and promotion mechanics are the
verify-companies skill, and the roster commit convention is in
CLAUDE.md. Scope a session to one provider — rosters land one per PR.
The providers in scope are the --provider values of
cmd/verify-companies (providerOrder). SmartRecruiters keeps a
roster too but has no adapter yet; spot-check its candidates with
go run ./cmd/smartrecruiters --company <id> search instead.
Sourcing channels
Work from the provider's public board-host shape — the entry for the
provider in careersHostPatternsByAdapter
(internal/ats/registry.go). Mix channels; each one alone dries up:
- Site dorks — WebSearch
site:<board host> <theme keywords>(e.g.site:jobs.lever.co fintech,site:job-boards.greenhouse.io robotics— for greenhouse the legacyboards.greenhouse.iohost is indexed too and redirects). Dork the public board host, never the API host — search engines index board pages, not API responses. Vary keywords by industry, region, and role; page past the first results. - Theme-first probing — start from a company list (industry
roundups, accelerator batches, "companies using " articles),
websearch
<company> careers, and keep the ones whose careers URL matches the provider's host shape. The slug comes from that URL, never from guessing the company name. - Migration finds — verify-companies audits surface companies that moved ATS; their new host names the provider and slug directly.
A careers URL matching a different supported provider's host shape is
still a find — park it in unverified/<other>.yaml rather than mixing
rosters in one session.
Candidate to entry
- Dedupe first:
grep -ri "<name-or-slug>" internal/provider/*/companies.yaml unverified/*.yaml.ats.NewRegistryonly requires unique display names and slugs within one adapter's own roster; a hit inside the target provider's file disqualifies or renames the candidate. A hit in a different provider's roster is fine — cross-adapter collisions are allowed andRegistry.Resolvedisambiguates them by careers URL at runtime, not a curation bug to dodge here. The one exception: when the target adapter cannot render a careers URL for the candidate, its display name is the only disambiguation key left, so that name must collide with no other entry's name or slug anywhere —NewRegistryfails otherwise. - Copy the YAML key shape from the target
companies.yaml— every provider names its identifier differently, and workday needs the tenant/instance/site triple from the careers URL, one row per tenant (the adapter dedupes by tenant and serves only the first). - Display name comes from an authoritative source (the company's own
site or the board's page title). A slug with an unconfirmed name
goes to
unverified/<provider>.yaml, not the curated roster.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2mo ago First seen · 100 lines · 82 tokens per session scan A 55dcca6aa435
discover-companies is a skill published in the GitHub repository amikai/openings-mcp (108 stars, last pushed 11d ago), licensed MIT. It adds 82 tokens to every session and 1,203 once invoked, about $0.0003 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
discover-companies compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| discover-companiesamikai/openings-mcp | Skill | 108 | 11d ago | MIT | A | 82 | Claude Code, Codex |
| speakblacktop/mcp-tts | Skill | 74 | 3d ago | MIT | A | 144 | Claude Code, Codex |
| live-auditGrinv/steam-games-mcp | Skill | 5 | yesterday | MIT | A | 83 | Claude Code, Codex |
| wikiTheWinci/mimirs | Skill | 30 | 18d ago | Apache-2.0 | A | 38 | Claude Code |
| prompt-checkGrinv/steam-games-mcp | Skill | 5 | yesterday | MIT | A | 54 | Claude Code, Codex |
| planTheWinci/mimirs | Skill | 30 | 18d ago | Apache-2.0 | A | 75 | Claude Code |
| researchTheWinci/mimirs | Skill | 30 | 18d ago | Apache-2.0 | A | 88 | Claude Code |