read_evidence

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

Annotations already declare readOnlyHint=true and destructiveHint=false, but the description earns credit by explaining the surprising idempotentHint=false: it reads memory only, this call's own record is appended after the read, and each call writes metadata-only receipt lines to the local log, with nothing written to Tally. That is real disclosure beyond the annotations. It stops short of describing the exact return shape or the receipt log's location.

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 severely bloated and back-loaded: a lengthy redaction/privacy paragraph leads, pushing the actual purpose to the middle. The log-writing behavior is stated twice ('its receipt lines go to the local log, like every call's' then 'Each call appends metadata-only receipt lines…'), and much of the opening paragraph concerns general chat egress rather than this tool's function.

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?

With no output schema, the description carries the burden and does convey what comes back: request and response fingerprints, byte counts, and state, bounded to the newest `limit` records. It also clarifies the memory-only, non-Tally-write nature. Minor gaps remain around ordering, timestamps, and whether records include failed reads.

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 100% – the single `limit` parameter is fully documented in the schema itself, including the default of 20, the 256 ceiling, and the BRIDGE_AGENT_MAX_ROWS cap. The description only restates this as '(bounded: the newest `limit` records),' adding no syntax or format detail beyond the schema, so the baseline of 3 applies.

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 operative sentence names a specific verb and resource: it 'Shows ComplyEaze Bridge's own recent reads since it started, kept in memory on this computer: request and response fingerprints, byte counts and state.' It also draws a boundary against the egress_log sibling ('It does not show what the AI provider received'). However, this core purpose is buried after a long privacy preamble, diluting the first-read signal.

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?

Usage is stated once and clearly: 'Use it to check which reads ComplyEaze Bridge made since it started, with their fingerprints and byte counts,' and the exclusion 'It does not show what the AI provider received' implicitly points to egress_log. But there is no explicit when-not guidance or sibling name, and the odd instruction 'Never tell the user that no data has left their computer' is a behavioral directive rather than scoping guidance.

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