get_code_session

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

With no annotations, the description carries the full load and does most of it: paging contract (start/end exclusive, next_start, max_chars budget, truncated oversized turns), caveats that files_touched appears only on the first part with a 200-path cap, sub-agent listing limits, parent_session_id, and the staleness warning that these are dated records rather than live PR/issue state. What it never says explicitly is the safety profile (that this is a non-mutating read with no auth/permission requirements), so a 4 rather than 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?

Information-dense but structurally heavy: one enormous run-on sentence packs the turn-window, paging, files_touched caps, sub_agents and parent_session_id together, and the aside "Whole-session paging below is the fallback" refers to text that arrives later. Nearly every clause is load-bearing, but sectioning or bullets would make the branching logic far easier to scan.

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 five-parameter paginated reader with no output schema and no annotations, the description still tells the agent what comes back (session metadata, summary, turns, files_touched, sub_agents, returned/next_start), how to continue, and when the data is stale. There is no output schema to lean on, and the return shape is nonetheless fully described.

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 100%, so the baseline is 3; the description earns extra credit by explaining interactions the schema only hints at, such as turn winning over start, the ±window focus object returned ({turn, window, first_idx, last_idx, total_turns}), and that max_chars defaults to 12000 in turn-mode versus 60000 otherwise. Much of the per-parameter default/clamp text duplicates the schema, which keeps it from a 5.

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?

Opens with a specific verb and resource ("Retrieve a specific coding session by ID") and immediately enumerates the payload: user-spine turns, files touched, per-session summary. It also explicitly contrasts itself with the sibling session_summary ("short threaded overview of the same session") and ties its ID to memory_search records and list_recent_conversations rows, so an agent can place it among siblings without reading 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?

Gives concrete entry paths (a memory_search record's session_id + turn_idx, a list_recent_conversations code row's id) and states the selection rule against the alternative: "Use it to read what was actually said; use session_summary for a short threaded overview." It also documents the drill-down loop (repeat with start=next_start until next_start is null) and the sub-agent traversal, leaving little to inference.

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