session_list

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

With no annotations provided, the description carries the full burden of safety disclosure, and it does this well: it states 'Read-only, makes no model call' and clarifies that it is 'Visibility-only — never affects oracle answers; oracles only see what a calling agent explicitly passes'. It also defines the internal staleness heuristic (~5 min heartbeat threshold) that shapes the `active_only` default. Minor deduction: it does not describe the exact output list ordering or whether entries are capped by `limit` before or after filtering, but those are less critical for selection and invocation.

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 dense but purposeful; every sentence adds behavioral or selection-relevant information. The opening word 'COORDINATION' acts as a category cue and front-loads the purpose. Minor deduction: the last sentence about visibility is slightly repetitive with the earlier 'Read-only, makes no model call' claim, so it could be tightened. Overall it is well under the length where an agent's attention starts to degrade.

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 zero-required-parameter list tool with no output schema, the description is nearly complete: it names the dashboard's purpose, what fields appear in entries, the defaults, the override flags, the staleness meaning, and the non-interference guarantee. The main gap is a lack of detail about pagination or truncation behavior (does `limit` default to 50 and cap at 200 in a simple head or does it page?), and it does not mention whether the ordering is by heartbeat recency. Still, an agent could confidently select and invoke it without further clarification.

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?

Schema description coverage is 67%: `active_only` and `all_projects` already have descriptions in the schema, while `limit` has only its default/min/max. The tool description adds the crucial semantics for the two booleans — e.g., that `active_only: false` yields 'retained history' and that `all_projects: true` means 'whole machine' — and it explains the default hint 'so the dashboard shows live work'. This is strong beyond-schema value. Deduct one point because the description does not add detail about `limit`'s effect beyond the schema's numeric bounds.

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 opens with a clear label ('COORDINATION — the operator dashboard') and states the specific verb+resource: lists ask_fable sessions across instances. It details exactly what each entry contains (session key, agent_id, latest question, oracle, status, heartbeat age, turn count), which strongly distinguishes it from siblings like session_stats, session_peek, and reset_session. An agent can immediately tell this is a read-only monitoring tool rather than a per-session query or mutation.

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 explicit defaults ('THIS project only', `active_only: true`) and explains when to flip them: pass `active_only: false` for retained history and `all_projects: true` for the whole machine. It also gives a concrete use case ('avoid duplicate work or watch the live fleet'), which helps an agent decide between this and session_peek or session_stats. This fully satisfies when-to-use guidance within the sibling set.

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