veezee-skills

Agent skills for Veezee: LinkedIn data for AI agents over MCP or REST, credit-metered, 500 free credits with no signup.

Each skill is a SKILL.md pack under skills/<name>/, in the standard agent-skills format (YAML frontmatter with name and description, body of instructions). Point your agent runtime at the pack, or copy it into your project's skills directory.

Skills

• prospect-enrich: enrich known prospects with current role, company, and experience data.

• candidate-sourcing: turn a role spec (title, company, school) into a ranked, enriched shortlist.

• company-enrichment: enrich a batch of companies (URLs, slugs, or website domains) with firmographics.

• account-research: one-account pre-outreach briefing: company facts, key people, recent posts.

• post-voice-research: what a person or company posts about: topics, tone, cadence.

Related MCP server: LinkedIn Api8 MCP Server

The server these skills drive

• MCP endpoint: https://mcp.veezee.io/mcp (streamable-http; legacy SSE at https://mcp.veezee.io/sse)

• Docs: https://veezee.io/docs, per-client setup: https://veezee.io/docs/clients

• Machine index of these skills: https://veezee.io/.well-known/agent-skills/index.json

Validation

npm test checks every pack: frontmatter shape, name matches its directory, and forbidden-vocabulary scan. CI runs it on every push.

Contributing

Truth rule: a skill may only instruct what the seven shipped tools actually do (see https://veezee.io/docs). Claims a test cannot verify do not merge.