Optional. The user's most recent messages, newest-relevant first.
source (optional) restricts to one platform label as saved (e.g.
"ChatGPT", "Claude"). limit optional (default 10, max 25).
| Name | Required | Description | Default |
|---|---|---|---|
| limit | No | ||
| source | No |
| Name | Required | Description | Default |
|---|---|---|---|
| error | No | ||
| inbox | No | ||
| message | No | ||
| instruction | No | ||
| upgrade_url | No | ||
| recent_messages | No |
Changes observed during successful MCP inspections. Dates show when Glama detected each change.
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnlyHint, idempotentHint, and destructiveHint, so the safety profile is covered. The description adds useful behavior beyond that: newest-relevant-first ordering, optional source filtering, and the limit default/max, all of which help an agent understand what to expect. Nothing in the description contradicts 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.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is short and front-loads the main purpose before parameter details. The leading 'Optional.' is slightly redundant and adds little, and the opening is a fragment, but overall it is tightly written with no significant wasted content.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a read-only tool with two optional parameters and an output schema, the description covers both parameters and the ordering behavior, which is sufficient for basic invocation. It does not explain what 'relevant' means or when the data is populated, but the low complexity and strong annotations keep this from being a serious gap.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The input schema has 0% description coverage, but the description fully compensates by explaining both parameters: source restricts to one saved platform label (with examples) and limit has an explicit default and maximum. This is exactly the semantic detail an agent needs beyond the bare schema field names.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description identifies the resource ('the user's most recent messages') and the ordering ('newest-relevant first'), which is clear enough for an agent to know what is returned. It lacks an explicit verb like 'get' or 'retrieve' and does not explicitly distinguish it from sibling search/context tools, so it is clear but not fully differentiated.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description gives parameter usage details such as source restricting to a platform label and limit defaulting to 10, but it provides no guidance on when to use this tool versus alternatives like search_semantic or get_context. There are no explicit scenarios, exclusions, or when-not-to-use instructions.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Add one secure layer between your agents and this server.
Most tools map cleanly to distinct resource+action pairs: projects, tasks, skills, user rules, memory, and messaging are all clearly separated. The main ambiguity is update_task versus update_task_state, since update_task can also change state and plan_status, though the descriptions do point to the narrow intended use.
The naming is largely consistent verb_noun snake_case: create_project, update_skill, delete_task, list_projects, get_context, save_turn. Minor deviations include recent_messages lacking a verb, remove_user_rule versus delete_* style, and singular user_rule in mutations versus plural user_rules in listing.
With 27 tools, the server is over the typical well-scoped MCP range, even though it covers several domains. Some consolidation is possible, such as folding update_task_state into update_task and reducing the overlapping retrieval/search tools.
The tool set provides strong lifecycle coverage for projects, tasks, skills, and user rules, plus memory retrieval, agent messaging, and onboarding help. Minor gaps exist, like no standalone get_task or list_tasks and no explicit inbox listing, but get_project and get_context largely cover those needs.
