Deadbolt

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

Annotations already declare readOnlyHint, idempotentHint and destructiveHint=false, so the safety profile is covered. The description adds genuinely new behavior: it never reads row data, it emits a single JSON value rather than a result set, and it must be executed externally via the Supabase SQL Editor or an execute_sql MCP tool. It stops short of noting practical execution limits (query runtime, result size, required privileges), which keeps it from a 5.

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?

It is front-loaded with the 'Use when' trigger and every clause carries a distinct fact (source of truth, what is read, what is excluded, return shape, how to run it, next tool). The first sentence is long and heavily parenthesized, which slightly hurts scannability, but there is no filler to cut.

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 zero parameters, no output schema and annotations covering the safety profile, the description supplies everything else an agent needs: why to choose it, what the SQL inspects and does not inspect, that the return is one JSON value, how to execute it, and where to send the result. Nothing material 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?

The tool takes zero parameters, so the baseline is 4. The description correctly implies there is nothing to configure — the SQL is generated as-is — and adds the meaningful invocation detail that execution happens outside this tool.

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 states a specific verb and resource ('returns a read-only SQL query') and enumerates exactly what the query covers: RLS settings, anon/authenticated grants, policies, SECURITY DEFINER functions, storage buckets. It also draws the key boundary against the sibling workflow ('rather than migration files'), so an agent can distinguish it without opening any schema.

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?

It gives an explicit trigger ('Use when you need to check a live Supabase project rather than migration files'), the alternative it replaces, and the follow-up routing ('pass the JSON to deadbolt_check_supabase_access'). The agent knows both when to pick this tool and what to do with its output.

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

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

Beyond the annotations (readOnly, idempotent, closed-world, non-destructive), it discloses the crucial behavioral facts: it is deterministic, never connects to any database, and enforces a 200,000-character input cap. It also enumerates the shape of the result (per-table anon/authenticated SELECT/INSERT/UPDATE/DELETE verdicts plus severity-ranked findings with fix SQL), which the annotations cannot convey.

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?

It is front-loaded with the triggering condition and organized around a single dense paragraph, with no filler sentences. The long enumeration of finding types and the parenthetical SQL keyword list make it heavier than necessary, but each element carries real information an agent would use.

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?

There is no output schema, yet the description fully compensates by describing the per-table verdict matrix and the severity/fix-SQL findings. For a nontrivial static-analysis tool, the agent has enough to decide to call it, choose the input form, and interpret what comes back.

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%, so both parameters are already fully documented in the input schema, including the two accepted input forms and the before/after Oct 30 2026 grant semantics. The description only echoes those forms and supplies the motivating context for the date flag, so it meets but does not exceed the baseline for a fully documented schema.

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 names a specific verb-and-resource pair: replaying Supabase/PostgreSQL migration SQL against real Postgres rules to determine per-table read/write access. It is clearly separable from the sibling deadbolt_catalog_query, which it explicitly consumes output from, and from unrelated siblings like doorman_check_robots_txt.

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?

It states exactly when to reach for it (writing/reviewing/applying migration SQL, user asking if data is exposed, diagnosing 42501 or empty results) and gives an explicit imperative: 'Run it before telling a user their schema is safe.' It also names the alternative input path (JSON from deadbolt_catalog_query) so the agent knows it need not re-run the catalog query.

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