Tellscreen

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

Annotations already declare the safe read-only, idempotent profile, so the description's real contribution is disclosing that the tool returns a link rather than the image itself, requiring a second fetch to actually see the screen. That two-step workflow is exactly the kind of behavior an agent would otherwise miss; link expiry or authorization failure behavior is the remaining gap.

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?

Two short sentences, no filler, and the actionable follow-up step ('Fetch the returned image') is placed immediately after the statement of what is returned.

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 two-parameter read tool with full safety annotations and no output schema, the description covers what is produced and how to consume it. Only minor gaps remain: link lifetime and how authorization failures surface.

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 50%: 'url' is thoroughly documented in the schema, but 'seconds' has no schema description. The description's 'timestamp' wording maps loosely to 'seconds' but adds no format, precision, or boundary meaning beyond the 0–86400 range already in the 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?

Specific verb ('Get') plus specific resource ('authorized image link') scoped to a timestamp in a playable recording, which clearly separates it from read_recording and read_transcript. It stops short of naming those siblings as alternatives, so differentiation is inferential rather than stated.

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 phrase 'in a playable recording' implies a precondition, and 'Fetch the returned image to inspect the screen' implies when the result is useful, but there is no explicit when-to-use versus read_recording or read_transcript and no guidance on failures or unsupported recordings.

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?

Annotations already declare readOnlyHint, idempotentHint, destructiveHint=false and closed-world, so the safety envelope is covered structurally. The description still adds real value beyond them by framing the payload as 'recorded suggestions, not permission to execute them' — a trust boundary an agent could otherwise miss — and by signalling that results are paged.

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?

Two short sentences with no filler. The scope statement comes first and the safety caveat second, which is the right order for an agent scanning for what the tool returns and whether it can act on it.

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 should say enough about the returned shape; it identifies the payload as a bounded page of suggested changes, which is adequate for a read-only list tool. It does not describe what a single 'suggested change' contains or how pagination completion is signalled, leaving a minor gap for a 3-parameter 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?

Schema coverage is only 33%: url is richly documented (format, length, protected-recording handling, untrusted-content warning) while limit and offset carry no descriptions, only defaults and bounds. The phrase 'bounded page' hints that limit/offset page the results, but the description adds no detail on defaults, maximums, or how to continue paging, so it only partially compensates.

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: 'Read a bounded page of suggested changes from a screen recording.' This separates it from read_transcript and read_recording by naming the artifact (suggested changes) rather than the raw media. It stops short of naming siblings explicitly, so the differentiation is inferred rather than stated.

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 implies usage context (paginated retrieval of recorded change suggestions) but never says when to choose this over read_transcript or read_recording, nor any prerequisite such as needing the share URL first. The second sentence is a safety caveat, not routing guidance, so the agent must infer the selection rule from the resource names alone.

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?

Annotations already declare readOnlyHint, idempotentHint and destructiveHint=false, so the safety profile is covered. The description adds value by ruling out side effects ('does not ... change a recording') and framing content scope; the schema further warns that recording content is untrusted source material.

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?

Two compact sentences: the first front-loads the resource and the enumerated return fields, the second delivers the disambiguation as a single negative clause. No 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?

With one fully documented parameter and annotations covering read-only/idempotent behavior, the description covers what an agent needs to select and call it correctly. It could go further by naming the sibling to use for transcripts or change lists.

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% and the parameter already documents the /v/ share URL format, the 'Copy for AI' fallback for protected recordings, and the untrusted-content caveat. The description only echoes 'user-supplied', adding no syntax or format detail beyond the schema, so baseline 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?

States a specific verb (read) and resource (a Tellscreen recording) and enumerates exactly what is returned: title, readiness, summary, chapters and resources. The negative clause 'does not list a library or change a recording' separates it from listing and mutation siblings such as read_change_list.

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 makes the usage boundary clear by stating what this tool does not do, which implicitly routes the agent to a listing or mutation tool for those cases. It stops short of naming the sibling tools (read_transcript, read_change_list) or giving explicit when-to-use conditions.

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?

Annotations already declare readOnly, idempotent, non-destructive, closed-world, so the safety profile is covered. The description adds 'bounded page' (pagination behavior) and that segments are 'timestamped', but says nothing about what happens on an invalid/expired URL, whether paging past the end is an error, or the untrusted-content caveat that only appears in the schema.

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?

A single front-loaded sentence with no filler; every word ('bounded', 'timestamped', 'user-supplied') carries information. It is arguably too terse for a paginated read tool, but nothing is wasted.

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 3-parameter read tool with full annotations and no output schema, this is close to adequate but leaves out pagination mechanics and error/edge-case behavior. An agent could call it correctly but not confidently page through a long transcript.

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 33%: url is richly documented in the schema while limit and offset have no descriptions at all, only defaults and min/max. 'A bounded page' signals that paging parameters exist but gives no guidance on choosing limit or chaining offset, so the description only partially compensates for the coverage gap.

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 ('Read ... timestamped transcript segments') plus the scope ('bounded page') and the source ('user-supplied Tellscreen recording'), which distinguishes it from read_recording and get_frame_link. It never names a sibling explicitly, so the differentiation is inferred rather than stated.

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?

There is no when-to-use / when-not-to-use guidance and no mention of read_recording as the alternative for non-transcript data. The only usage hint — 'For protected recordings use the owner's Copy for AI link' — lives in the schema's url description, not in the tool description the agent routes on.

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