MCP Recruiting Agent
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| ANTHROPIC_API_KEY | No | Your Anthropic API key. Required when using the Anthropic LLM (RECRUITING_AGENT_LLM=anthropic). | |
| RECRUITING_AGENT_LLM | No | Language model to use. Set to 'anthropic' to run with Claude. Defaults to the offline deterministic model. | offline |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": false
} |
| prompts | {
"listChanged": false
} |
| resources | {
"subscribe": false,
"listChanged": false
} |
| experimental | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| search_jobsA | Search open jobs by keywords (skills or title words) and optional location. |
| get_jobC | Get one job posting by id, e.g. J-101. |
| search_candidatesB | Rank candidates by overlap with the given skills. Never returns contact details. |
| get_candidateA | Get one candidate profile by id, e.g. C-201. Contact details are excluded. |
| score_matchC | Transparent skill/experience match score between a candidate and a job. |
| get_policyA | Get hiring policy text. Topics: eeo, screening, data_privacy, ai_use. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
| shortlist | Prompt template: build a skills-only shortlist for a job. |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 6 tools
Each tool targets a distinct action: searching vs retrieving jobs/candidates, matching a candidate to a job, and fetching policy. There is no overlap in purpose; even search vs get are clearly separated by scope (list vs single record).
All tool names follow a consistent verb_noun pattern with snake_case (e.g., search_jobs, get_candidate, score_match). The verbs are action-oriented and the nouns are domain objects, making the set predictable and easy to navigate.
With 6 tools, the server is well-scoped for a recruiting assistant: two search tools, two retrieval tools, a matching tool, and a policy lookup. Each tool serves a clear purpose without redundancy or bloat.
The surface covers the core recruiting workflow of discovering jobs and candidates, examining details, scoring matches, and consulting hiring policies. Since this appears to be a read-only analysis agent, no update/create/delete operations are required; the set has no obvious dead ends.