Skip to main content
Glama
Sign In

Jam

by dev.jam
OverviewDiscussionsAdmin

submitFeedback

Save feedback about a Jam surface in Jam's database for Jam's agents to read. Returns the stored feedback ID and a thank-you. Use it when a Jam MCP tool, CLI command, doc page, or skill was wrong, stale, confusing, or missing something — and when one worked unusually well. Also send it when you find the cause of a Jam's problem or fix it: set jamId to that Jam and evidence to the Jam data that led you there, and say in the body what in it pointed to the cause. Report problems only with Jam's own surfaces: not with another MCP server's tools, the user's own skills or scripts, or the user's product, because Jam cannot fix those. The evidence report is for bugs in the user's product. Call it yourself; do not ask the user for approval, because no person reads it. Send at most one problem report per session, plus one evidence report for each Jam whose cause you find, and never for a slip you corrected yourself. Keep the body within 2000 words and 100,000 characters, and specific enough to act on: what you were doing, what you expected, what happened, and the mistake. Larger bodies are rejected without saving or truncating them. Never include secrets, customer data, source code, or stack traces — describe code rather than pasting it.

Input Schema

NameRequiredDescriptionDefault
bodyYesThe feedback itself, at most 2000 words and 100,000 characters. For a failure: what you were doing, what you expected, what happened, and the mistake.
jamIdNoThe Jam whose cause you found, when you send evidence — a UUID or a jam.dev URL
intentNoIn 15-30 English words, your larger task and why this call helps it. People by role ("a customer"); no argument values or secrets.
subjectNoThe exact thing: MCP tool name, CLI command path, doc URL, or skill name
categoryNoWhat the feedback is about. Defaults to unknown.
evidenceNoWhen you found the cause or fixed the bug: the Jam data that led you there, in the order it mattered: first the source without which you would not have found the cause. description is the title and reporter's text, transcript the voiceover, user_actions the clicks, typing and navigations, and metadata the browser, device and custom metadata. Only data you read in this session.
llm_modelNoYour exact model id from your system prompt, or "unknown". Never guess.

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed2 schema fields changed
    • changedInput schema / properties / intent / description
      Previous value: -"In 15-30 words, say what larger task you are working on and why this tool helps it. Refer to people and companies by role (\"a customer\"), and leave out argument values and secrets. Use English. Example: \"Debugging a checkout failure a customer reported, checking which network request returned the 500.\""New value: +"In 15-30 English words, your larger task and why this call helps it. People by role (\"a customer\"); no argument values or secrets."
    • changedInput schema / properties / llm_model / description
      Previous value: -"The exact model identifier you are running as, from your system prompt or environment (e.g. \"claude-opus-4-8\", \"gpt-5.2\"). Pass \"unknown\" if you are not certain; never guess."New value: +"Your exact model id from your system prompt, or \"unknown\". Never guess."
  2. First observed

TDQS

A4.8/5.0
Behavior5/5

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

Goes well beyond the annotations: it discloses that no human reads the report and no approval is needed, that oversized bodies are rejected without saving or truncating, the word/character ceilings, and the prohibition on secrets, customer data, source code, and stack traces. These are operational facts an agent cannot infer from readOnlyHint=false / destructiveHint=false alone.

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 return value, then routing rules, then limits. It is dense but almost every clause changes agent behavior; the only real redundancy is restating the 2000-word/100,000-character ceiling already present in the `body` schema description.

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 a 7-parameter mutation with no output schema, the description covers what is written, what is returned, who consumes it, when to send it, and how often — enough to call it correctly on the first attempt without further context.

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

Parameters4/5

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

Schema coverage is 100%, so the baseline is 3; the description adds real meaning on top by explaining the jamId + evidence pairing for cause reports and the ordering rule for `evidence` ("first the source without which you would not have found the cause"). It does not further explain `intent`, `category`, or `llm_model`, which the schema already covers.

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: "Save feedback about a Jam surface in Jam's database for Jam's agents to read," plus the return value (stored feedback ID and a thank-you). No sibling tool covers feedback submission, and the description makes the tool's distinctive role unmistakable.

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 when-to-use (tool/CLI/doc/skill was wrong, stale, confusing, missing — or worked unusually well; or you found a Jam's root cause), explicit when-not (not another MCP server's tools, not the user's own skills/scripts, not the user's product, not slips you corrected yourself), and explicit rate limits (one problem report per session plus one evidence report per Jam).

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

All Jam toolsInspect
Try in Browser
Claude Code, inside the Godot 4 editorFlockbay puts an AI agent in the real Godot 4 editor: it writes GDScript, runs your game and fixes what breaks. Free, or bring your own Claude Code or Codex.flockbay.comAd

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources