Cortex

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

With no annotations provided, the description carries the full burden of behavioral disclosure. While 'save/restore' implies both read and write capabilities, it doesn't specify whether this is destructive, whether it requires special permissions, what the persistence characteristics are, or what happens during failures. The metaphorical language ('hippocampal replay', 'context compaction events') further obscures the actual behavior rather than clarifying it.

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 concise (one sentence) but inefficiently uses that space. The metaphorical framing ('hippocampal replay', 'context compaction events') adds complexity without clear benefit. While brief, it's not front-loaded with actionable information - the core function 'save/restore working state' is buried in metaphorical language rather than stated clearly upfront.

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 tool with 10 parameters (1 required, 9 optional), 0% schema description coverage, no annotations, but with an output schema, the description is inadequate. It doesn't explain what the tool actually does in concrete terms, provides no parameter guidance, and relies on metaphorical language that obscures rather than clarifies. While the output schema might document return values, the description fails to provide the necessary context for understanding when and how to use this complex state management tool.

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?

With 10 parameters and 0% schema description coverage, the description provides no information about any parameters. It doesn't mention the required 'action' parameter or any of the 9 optional parameters like 'directory', 'current_task', 'files_being_edited', etc. The description fails to compensate for the complete lack of schema documentation, leaving all parameters semantically unexplained.

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 'Hippocampal replay: save/restore working state across context compaction events' uses metaphorical language ('hippocampal replay', 'context compaction events') that makes the purpose somewhat vague. It mentions 'save/restore working state' which indicates a state management function, but the metaphorical framing obscures the concrete action. It doesn't clearly distinguish this from sibling tools like 'record_session_end', 'remember', or 'backfill_memories' which might also involve state management.

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 provides no explicit guidance on when to use this tool versus alternatives. The phrase 'across context compaction events' hints at a specific triggering condition, but doesn't explain what those events are or when they occur. There's no mention of prerequisites, timing considerations, or comparison to sibling tools like 'remember' or 'record_session_end' that might serve similar purposes.

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