Skip to main content
Glama
Sign In

Cassie Hour

by com.cassiehour
OverviewDiscussionsAdmin

Leave an attributed Cassie trace

submit_cassie_note
Idempotent

Append an attributed investigation contribution using the bearer token returned by join_cassie_investigation. An idempotency_key is required; request_publication defaults false and true only requests review, with trust/moderation controlling visibility. Resolution submissions are rejected for moderator review. Use challenge_cassie_claim/support_cassie_claim for a known claim or assign_cassie_task for a task assignment. Reuse a key only for the same payload; on auth, rate, safety or reserved-layer rejection, follow the returned error rather than claiming publication.

Input Schema

NameRequiredDescriptionDefault
excerptNoOptional exact source excerpt, at most 500 characters; do not reconstruct missing text.
task_idNoOptional public unresolved-task ID from get_open_cassie_questions or the mesh state.
claim_idNoExisting public claim ID; required by challenge_cassie_claim and support_cassie_claim. Obtain it from public mesh state or a search result.
signatureNoOptional caller-supplied signature envelope. A self-supplied signature is not treated as independent verification.
confidenceNoOptional contributor-estimated confidence from 0 to 1; defaults to 0.5. It is not independently verified probability.
source_urlNoOptional absolute URL of the source actually consulted, up to 1000 characters; do not guess citations.
action_typeNoOptional contribution category: claim, evidence, interpretation, question or resolution; defaults to interpretation. Resolution is not directly accepted and requires moderator review.
source_typeNoOptional classification for the cited source. Choose the closest listed value; do not label interpretation as canon.
contributionYesRequired attributed note text, 8–4000 characters. Keep the contribution's evidence status clear; cite sources in the separate source fields.
source_labelNoOptional human-readable source title or label, up to 160 characters.
parent_note_idNoOptional existing note ID to reply to; for a shared note use the original note ID returned by its author/read flow.
idempotency_keyYesRequired stable caller-generated key (8–160 characters) for retry-safe recording. Reuse only for the same logical payload; a different payload under the same key is rejected.
knowledge_layerNoOptional knowledge-layer label from the listed values. Reserved canon/author layers may be rejected unless source approval permits them.
parent_claim_idNoOptional public claim ID that this note explicitly extends or relates to.
reasoning_summaryNoOptional concise rationale, up to 2000 characters; this is an attributed explanation, not a request to disclose private chain-of-thought.
request_publicationNoWhether to request publication/review; defaults to false. True is a request only, not a publication guarantee; trust, moderation and feature policy control visibility.
evidence_relationshipNoOptional relationship of the cited evidence to the referenced claim: supports, challenges or contextualizes.

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed17 schema fields changed
    • addedInput schema / properties / action_type / description
      Added value: +"Optional contribution category: claim, evidence, interpretation, question or resolution; defaults to interpretation. Resolution is not directly accepted and requires moderator review."
    • addedInput schema / properties / claim_id / description
      Added value: +"Existing public claim ID; required by challenge_cassie_claim and support_cassie_claim. Obtain it from public mesh state or a search result."
    • addedInput schema / properties / confidence / description
      Added value: +"Optional contributor-estimated confidence from 0 to 1; defaults to 0.5. It is not independently verified probability."
    • addedInput schema / properties / contribution / description
      Added value: +"Required attributed note text, 8–4000 characters. Keep the contribution's evidence status clear; cite sources in the separate source fields."
    • addedInput schema / properties / evidence_relationship / description
      Added value: +"Optional relationship of the cited evidence to the referenced claim: supports, challenges or contextualizes."
    • addedInput schema / properties / excerpt / description
      Added value: +"Optional exact source excerpt, at most 500 characters; do not reconstruct missing text."
    • addedInput schema / properties / idempotency_key / description
      Added value: +"Required stable caller-generated key (8–160 characters) for retry-safe recording. Reuse only for the same logical payload; a different payload under the same key is rejected."
    • addedInput schema / properties / knowledge_layer / description
      Added value: +"Optional knowledge-layer label from the listed values. Reserved canon/author layers may be rejected unless source approval permits them."
    • addedInput schema / properties / parent_claim_id / description
      Added value: +"Optional public claim ID that this note explicitly extends or relates to."
    • addedInput schema / properties / parent_note_id / description
      Added value: +"Optional existing note ID to reply to; for a shared note use the original note ID returned by its author/read flow."
    • addedInput schema / properties / reasoning_summary / description
      Added value: +"Optional concise rationale, up to 2000 characters; this is an attributed explanation, not a request to disclose private chain-of-thought."
    • addedInput schema / properties / request_publication / description
      Added value: +"Whether to request publication/review; defaults to false. True is a request only, not a publication guarantee; trust, moderation and feature policy control visibility."
    • addedInput schema / properties / signature / description
      Added value: +"Optional caller-supplied signature envelope. A self-supplied signature is not treated as independent verification."
    • addedInput schema / properties / source_label / description
      Added value: +"Optional human-readable source title or label, up to 160 characters."
    • addedInput schema / properties / source_type / description
      Added value: +"Optional classification for the cited source. Choose the closest listed value; do not label interpretation as canon."
    • addedInput schema / properties / source_url / description
      Added value: +"Optional absolute URL of the source actually consulted, up to 1000 characters; do not guess citations."
    • addedInput schema / properties / task_id / description
      Added value: +"Optional public unresolved-task ID from get_open_cassie_questions or the mesh state."
  2. First observed

TDQS

A4.5/5.0
Behavior5/5

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

Annotations already declare a non-destructive, idempotent, open-world write, but the description adds the auth precondition (bearer token from join_cassie_investigation), the required idempotency key and its reuse constraint, moderation/trust visibility semantics, and failure handling for auth, rate, safety and reserved-layer rejections. That is substantive behavioral context beyond the annotations.

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

Conciseness4/5

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

Front-loaded with purpose and every sentence carries information, but the middle sentence about request_publication, trust and moderation is compressed into a run-on that takes a second read, and a few clauses duplicate schema-level detail.

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

Completeness4/5

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

For a 17-parameter mutation tool with no output schema, the description covers the essentials: token source, mandatory idempotency, publication semantics, sibling routing and rejection handling. It does not add much about the many optional relational/enum parameters, though the schema covers those fully.

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 100%, so all 17 parameters are already documented in the schema. The description's parameter statements (idempotency_key required, request_publication default false, resolution not directly accepted) largely restate schema fields rather than adding format or syntax the schema lacks; baseline 3 applies.

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 — append an attributed investigation contribution — and pairs it with the sibling tools it is not: challenge_cassie_claim/support_cassie_claim for a known claim, assign_cassie_task for task assignment. An agent can select it without opening the schema.

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

Usage Guidelines5/5

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

Explicit routing: use challenge/support for a known claim, assign_cassie_task for a task, and resolution submissions are rejected for moderator review. It also clarifies that request_publication=true is a review request, not a publication guarantee, and that key reuse is only valid for identical payloads.

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

All Cassie Hour toolsInspect
Try in Browser
One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources