# ResearchPilot

> Plan and verify evidence-aware research workflows with specialist execution receipts.

- name: io.github.cnucho/researchpilot
- version: 0.2.0
- connection class: R0 (autonomous)
- trust: 94/100
- quality: 91/100 (strong)
- owner: not claimed
- descriptor: https://api.protogrid.dev/v1/servers/io.github.cnucho%2Fresearchpilot
- tools: https://api.protogrid.dev/v1/servers/io.github.cnucho%2Fresearchpilot/tools

**An agent can connect right now, no human step.** Remote endpoint, no authentication, reachable on the last probe.

## Connection (mcpServers)

```json
{
  "mcpServers": {
    "researchpilot": {
      "type": "http",
      "url": "https://researchpilot-academy-production.up.railway.app/mcp"
    }
  }
}
```


## Trust

- hygiene: 100
- liveness: 100
- freshness: 100
- provenance: 80
- drivers: +repository +namespace-matches-repo +reachable +uptime-100% +updated-15d-ago
- Derived from observable signals (official registry feed and our own credential-free probes); no code audit performed.

## Quality 91/100 (strong)

### protocol: 75 (weight 25)

- warn `protocol.modern`: newest supported version is 2025-11-25; 2026-07-28 not supported, older versions are deprecated until 2027-07-28
- n/a `protocol.stateless`: only applies to 2026-07-28 servers
- pass `protocol.transport`: streamable HTTP
- n/a `protocol.list_ttl`: only applies to 2026-07-28 servers

### authorization: n/a (weight 15)

- n/a `auth.prm`, `auth.as_metadata`, `auth.cimd`: no remote uses OAuth
- n/a `auth.secret_in_url`: no templated URL

### tool hygiene: 100 (weight 30)

- pass `tools.descriptions`: every tool has a description
- pass `tools.description_length`: descriptions are concise
- pass `tools.schemas`: every tool has an object input schema
- pass `tools.annotations`: 23 of 23 tools declare readOnlyHint or destructiveHint
- pass `tools.directory_hints`: every tool declares readOnlyHint, destructiveHint, idempotentHint and openWorldHint
- pass `tools.token_cost`: about 3,762 tokens to load every tool
- pass `tools.api_dump`: tools are not a one-to-one API dump

### stability: 100 (weight 15)

- pass `stability.changes`: no tool changes in 30 days
- pass `stability.rug_pull`: no tool changed its meaning under the same name

### dependencies: n/a (weight 15)

- n/a `deps.known_vulns`, `deps.mcp_sdk_version`, `deps.resolvable`: no npm or PyPI package

- tools: 23, about 3,762 tokens to load them all
- tool set hash: ce53159aab800561e78c2e8a92562029
- badge: [![protogrid quality](https://protogrid.dev/badge/io.github.cnucho/researchpilot.svg)](https://protogrid.dev/servers/io.github.cnucho/researchpilot)
- Checks run on what our credential-free, read-only probes observe; tools are never called and no code audit is performed.

### Tool changes

_No tool definition changes recorded._

## Remotes

- https://researchpilot-academy-production.up.railway.app/mcp (streamable-http, auth none, reachable true, uptime30d 1)

## Packages

_none_

## Tools (23)

- `researchpilot_build_collection_recipe`: Build a reproducible collection plan for YouTube metadata, web text, legal references, data APIs, or a bounded text corpus. This returns a plan only and never claims collection occurred.
- `researchpilot_capabilities`: Read canonical ResearchPilot capability authority. Use this before claiming that ResearchPilot can directly execute a research method.
- `researchpilot_check_method`: Check one ResearchPilot capability by canonical capability ID and distinguish discussion/planning, direct execution, certified core, registered engine, and specialist handoff.
- `researchpilot_content_analysis_template`: Create a bounded content-analysis codebook and validation skeleton with sampling/coding/context units, category contracts, pilot coding and human/AI validation requirements. Categories remain researcher proposals.
- `researchpilot_get_profile`: Return the authenticated ResearchPilot user profile for the current OAuth connection. Use only for user identity/personalization, not as research evidence.
- `researchpilot_get_project_context`: Read one authenticated user's cloud ResearchPilot project state: research question, decisions, analysis/result summaries, stale state, source identities/digests and local-artifact availability. Raw local research files are intentionally absent.
- `researchpilot_get_user_memory`: Read the authenticated user's ResearchPilot working preferences and bounded routing history. These are workflow/presentation signals, never scientific authority.
- `researchpilot_handoff_context`: Read the immutable ResearchContext snapshot carried by a ResearchPilot handoff, including an evidence catalog without source bodies. This does not mutate Studio state or authorize execution.
- `researchpilot_handoff_summary`: Validate the outer ResearchPilot specialist handoff identity and return a bounded context summary without source bodies. Use the original handoff package as the authority.
- `researchpilot_list_projects`: List the authenticated user's active ResearchPilot projects, newest first. Use before choosing a project when the user's phrase such as 'my study' is ambiguous.
- `researchpilot_list_work_artifacts`: List saved working artifacts for one authenticated ResearchPilot project. Use this to continue prior host-AI plans, interpretations, critiques, or drafts without treating them as researcher-approved canonical decisions.
- `researchpilot_open_workspace`: Open the interactive ResearchPilot workspace when the user wants to select a project, inspect method capability, review research context, or edit an instruction before continuing in the host AI conversation.
- `researchpilot_prepare_specialist_execution`: Prepare a rigorous specialist execution request bound to the authenticated project, exact ResearchContext digest, capability authority, and input artifact digests. This does not execute the analysis or transfer local raw files.
- `researchpilot_read_handoff_evidence`: Read one exact evidence page from a ResearchPilot handoff after verifying the outer source digest against the immutable ResearchContext snapshot.
- `researchpilot_read_work_artifact`: Read one saved working artifact after verifying that both project and artifact belong to the authenticated ResearchPilot owner.
- `researchpilot_record_routing_feedback`: Record explicit user feedback about a ResearchPilot capability/tool outcome. Call only when the user explicitly accepts, rejects, corrects, or reports failure/success. Do not store the user's full research question or source text.
- `researchpilot_research_registry`: Search ResearchPilot's unified method/model, source-provider, official-reference, or planned-plugin registry. Planning/discovery only; registry presence never authorizes execution or collection.
- `researchpilot_save_work_artifact`: Save a substantive project-specific plan, interpretation, critique, draft, or specialist result as a non-canonical WORKING artifact. Use for durable ResearchPilot continuity; this never accepts a research decision or authorizes execution.
- `researchpilot_search_datasets`: Search ResearchPilot dataset/data-paper providers with explicit provider availability states. Returned records remain source candidates, not automatic dataset-fit approval.
- `researchpilot_search_literature`: Search research literature through ResearchPilot's OpenAlex/Crossref adapter. Bibliographic verification is not scientific-quality or full-text verification.
- `researchpilot_text_analysis_pilot`: Create a bounded text-analysis pilot package: collection recipe, codebook skeleton, corpus manifest, pilot/validation plan, AG-UI state, and specialist handoff route. Design only; no coding or validation is claimed.
- `researchpilot_update_user_preferences`: Update bounded ResearchPilot working/writing preferences only when the user explicitly states a stable preference. These preferences affect presentation/workflow, never scientific authority or method validity.
- `researchpilot_validate_specialist_receipt`: Validate a specialist execution receipt against its ResearchPilot request and authenticated project/context identity. This verifies identity, input digests, sample accounting, diagnostics and validation-state structure; it does not independently validate scientific correctness.