Skip to main content
Glama
StepenkoAnatoli

Fetchproof

by StepenkoAnatoli
OverviewSchemaSourceRelated ServersDiscussionsAdmin
JavaScript
Local

Collect research evidence

collect

Starts a GitHub Actions research run and returns a workflow run ID immediately; use that ID with fetch_corpus to retrieve the unapproved corpus for review.

Instructions

Start a research collection on GitHub Actions. Returns the workflow run id immediately; the run takes minutes. Call fetch_corpus with that id to get the result. The corpus that comes back is NOT approved research: the agent must review it before anything is built from it.

Input Schema

NameRequiredDescriptionDefault
urlsNoOptional. Pages you already know (http or https), fetched directly instead of searched - an API response, a changelog post. Each counts against max_pages. With urls and no queries, nothing is searched. Visible to anyone who can read that repository, like the topic: never a signed or token-bearing URL.
depthNoCollection tier. Default quick.
priorNoOptional. What you EXPECT the evidence to say, and what you know you cannot know yet. Registered on the runner and chained ahead of the first page, so it can only be supplied now. Nothing grades it - being wrong is the point (ADR-0039).
topicYesWhat to research. Visible to anyone who can read that repository.
preferNoOptional, comma-separated. Domains that OWN the fact - e.g. "tavily.com". On a shared host name the path - "github.com/actions/upload-artifact", not "github.com". Ranked above pages merely about it.
runnerNo
queriesNoOptional. The actual search queries. Without them the topic is used verbatim, which matches the words rather than the subject when the topic is made of common ones.
max_pagesNoPages to collect. Each costs at least one credit. Default 8.
client_refNoOptional caller reference. PUBLIC: it becomes the run and artifact name, so never put the subject in it.
repositoryYesowner/name of the repository holding the collector workflow

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observedv0.1.0

TDQS

A4.3/5.0
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description carries the full burden and does well: it discloses the asynchronous behavior, the latency ('the run takes minutes'), and the trust/quality caveat on the output. It does not cover failure modes, concurrency limits, or credit consumption at the description level, which keeps it below a 5.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Three short sentences, front-loaded with the action and the return value, then the follow-up, then the caveat. Zero filler and no repetition.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For an async launcher with rich schema coverage and no output schema, the description supplies exactly the missing pieces: that the call returns an id rather than results, that the run is slow, and how to retrieve and treat the result. An agent has everything needed to invoke it correctly.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 90%, so parameters like urls, prior, prefer, and client_ref are already well documented in the schema. The description adds no parameter-level meaning beyond the schema; baseline 3 applies when the schema does the heavy lifting.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb and resource ('Start a research collection on GitHub Actions') and immediately clarifies the async contract: it returns a workflow run id, not results. This cleanly separates it from the sibling fetch_corpus, which consumes that id.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Explicitly names the follow-up step ('Call fetch_corpus with that id to get the result') and sets a hard usage constraint: the returned corpus is not approved research and must be reviewed before anything is built from it. Lacks guidance on when not to use it (e.g. reusing an existing run), but the forward path is unambiguous.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Deploy Server

Other Tools