Parallect MCP Server
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| PARALLECT_API_KEY | Yes | Your Parallect API key (par_live_...) | |
| PARALLECT_API_URL | No | API base URL (for self-hosted or dev) | https://parallect.ai |
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": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| researchA | Submit a deep research query. Always creates a new research thread. To follow up on existing research, use the follow_up tool instead. Fans out to multiple AI research providers and synthesizes results into a unified report with cross-referenced citations and conflict resolution. |
| research_statusA | Check the status of a research job. Returns progress for each provider and overall job status. |
| get_resultsA | Get the synthesized results of a completed research job. Returns the unified report with [src_N] inline citations, a sourceRegistry to resolve those citations to URLs, follow-on suggestions, and cost breakdown. Set includeClaimsJson to get first-class claims with source links and confidence scores. |
| follow_upB | Pursue a follow-on research question from a completed job's suggestions, or ask a custom follow-up in the same thread. |
| list_threadsC | List recent research threads for the authenticated user. |
| get_threadA | Get a specific research thread with all messages, jobs, and results. |
| balanceB | Check current credit balance and optionally view recent transactions. |
| usageB | Get usage analytics: spend by provider, job count, average cost per query. |
| list_providersA | List available research providers, their models, capabilities, and the default provider assignments per budget tier. |
| search_claimsB | Search for claims across research jobs. Can search by text content, filter by confidence level, or list all claims for a thread. |
| get_claim_evidenceB | Get the full evidence chain for a specific claim, including all supporting sources, their URLs, snippets, and the claim's history of events. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 11 tools
Tools target distinct stages of the research lifecycle and the descriptions explicitly resolve the riskiest overlap (research creates a new thread vs follow_up continues an existing one; research_status vs get_results distinguish progress from output). The only mild overlap is usage (spend analytics) versus balance (credit balance), but both descriptions make the boundary clear.
All names are snake_case, but verb styles are mixed: some are verb_noun (get_results, list_threads, search_claims, get_claim_evidence), some are bare nouns (research, usage, balance), and one is noun_status (research_status). It remains readable, but the pattern is not predictable.
Eleven tools sit squarely in the well-scoped 3-15 range. Each tool maps to a distinct capability (submit, poll, retrieve, follow up, browse threads, inspect claims, list providers, view spend/balance) with no filler.
The surface covers the full research lifecycle: submission, status polling, result retrieval, follow-ups, thread history, claim/evidence inspection, provider listing, and usage/billing. Minor gaps exist (no explicit cancel-job or delete-thread operation), but agents can work around these.