propose_action

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

With no annotations, the description carries the full burden and does so: it declares the tool is non-executing, that params are validated at proposal time, that confidence is persisted with the proposal, that writeback targets are allow-listed and SSRF-policy-checked, and that dry_run sends nothing. The 401-on-Supabase disclosure is exactly the kind of auth behavior an agent cannot infer from any structured field.

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?

Front-loaded with the core behavior and the negation, and every paragraph adds real information. The multi-user/Supabase paragraph is dense and slightly tangential to invoking the tool, and the hard line breaks mid-sentence make it less scannable than it could be, but nothing is filler.

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?

An output schema exists, so return values need no explanation; the description instead covers the things structured fields omit — execution model, validation timing, dry-run prerequisite, and deployment-specific auth failures. Nothing an agent needs to call this correctly is missing.

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 coverage is 0% — no property has a description and `name` has no enum — so the description must compensate, and it does: `name` is enumerated with all five registered action values, `params` is defined as that action's own param object with validation semantics, and `confidence` gets both its 0-1 range and its meaning (self-reported, recorded with the proposal). This is substantially more than the schema provides.

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?

States a specific verb and resource — queue a governed action proposal for operator approval — and immediately negates the most likely misreading ("This does NOT run the action"). It also names the concrete action set (flag_entity, promote_incident, nominate_target, add_watch, writeback), which separates it cleanly from approve_proposal/reject_proposal/list_proposals.

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?

Gives strong when-to-use context (propose now, operator approves later; params validated here rather than at approval), a mandatory prerequisite for the writeback path (`dry_run: true` first), and an explicit environment exclusion (401 by design on Supabase multi-user deployments). What it never addresses is the apparent overlap with the sibling `create_watch_rule` versus proposing `add_watch` — an agent may hesitate between those two.

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