Base of AGI

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

Annotations already establish read-only, idempotent and non-destructive behavior, but the description adds substantial context beyond them: 'Rule-based static check: no LLM, nothing is executed' and 'Evidence is untrusted text from the repository.' It also enumerates the return payload (SPDX license, commercial_use semantics, risk flags with file/line, verdict), which is exactly the behavioral disclosure an agent needs.

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 purpose and input, then output, then the execution-safety caveat. Despite being dense, every clause carries information (accepted inputs, return fields, static-check semantics) with 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 no output schema, the description compensates by listing the returned fields (license, commercial_use, risk flags with location, last push, stars, verdict). Combined with the input-format guidance and the untrusted-evidence warning, an agent has everything needed to call and interpret this 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 100%, so the baseline is 3, but the description genuinely extends the schema by enumerating the accepted URL forms (SKILL.md file, skill folder, repository, or raw.githubusercontent.com file URL), clarifying the 'url' parameter beyond the schema's example. The 'lang' parameter is left to 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?

States a specific verb and resource ('License and risk report for an agent skill on GitHub') plus the scope of what it reports. An agent immediately knows this is a pre-installation static audit, not a general GitHub lookup.

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 'before the user installs it' clearly establishes the trigger context, and the input guidance ('Give a github.com URL of a SKILL.md file, a skill folder or a repository') tells the agent what to supply. It does not name explicit when-not conditions, but no sibling tool overlaps this purpose, so the omission is minor.

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 and openWorld=false, so safety is covered. The description adds real behavioral context not in the schema: rule-based tagging and a 30-day freshness window (plus the 3-day window when tags are empty, which lives in the schema). It does not mention rate limits, caching, or latency.

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 sentences, front-loaded with what the tool returns and its tag taxonomy, then invocation and the 30-day constraint. Dense with zero filler, though the parenthetical "(rule-based)" and the window note could be folded more tightly.

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 6-parameter, all-optional read tool with no output schema, the description covers what is returned and how to scope it. It omits any hint of the response shape (grouping by tag or type, ordering) and how `limit` interacts with the window, which are minor gaps.

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 83%, so the schema already documents lang, tags, since and exclude_ids well. The description reinforces the tags parameter with concrete examples and notes since is optional and the overall data window, but adds little beyond the schema's own parameter text. Baseline 3 is appropriate.

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 concrete resource (a developer's stack brief) and enumerates its content types (security notes, breaking changes, model releases, new tools, news), so the agent knows exactly what it returns. It distinguishes itself implicitly from the news-only siblings by tying items to stack tags, but never names latest_news or search_news to draw the line explicitly.

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?

"Pass the stack tags ... and optionally a since date" gives invocation guidance for this tool but offers no when-to-use versus the sibling news tools (latest_news, search_news, find_tools). Usage is implied by the stack-tag framing rather than stated as a routing rule.

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 and closed-world, so safety is covered. The description adds value beyond that by disclosing what each result carries (license, commercial_use derived from SPDX, install info as the source states it, source repo) and by setting a behavioral constraint against installing without approval — valuable since there is no output 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?

Two tightly packed sentences with no filler: the first defines the corpus searched, the second defines the result payload and the approval constraint. It is front-loaded with the action and leaves nothing redundant.

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 responsibly explains what each result contains, which is the main gap a caller would otherwise have. It stops short of covering the lang filter's skip behavior or limit semantics, but the schema handles those, so the definition is essentially complete for a 5-param search 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 description coverage is 80%, so the schema already documents query, kind, lang and commercial_use. The description's only semantic addition is that commercial_use derives from the SPDX license, which is largely repeated in 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 ('Search') and enumerates the resource types covered (MCP servers, Claude Code plugins, skills, hand-picked agent tools), which is far more informative than the title. It implicitly separates discovery from the sibling get_tool's retrieval role, but never names that sibling to make the distinction explicit.

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?

'Suggest tools to the user; never install without approval' gives workflow guidance on how to act on results, which is useful. However, there is no stated condition for choosing this over get_tool or check_skill, so the when-to-use-vs-alternative dimension is only implied by the word 'Search'.

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 cover the safety profile (readOnly, idempotent, non-destructive, closed-world), so the description's real contribution is disclosing the return payload: license, commercial_use, install, repository, stars, last push, preinstalled surfaces. The caveat 'as the source states it' usefully signals that install data is passed through unnormalized and may be unreliable — genuine behavioral context beyond structured fields.

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?

One sentence, front-loaded with the resource and its provenance, then a tight field list. Nothing is padded and no sentence fails to earn its place.

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 carries the burden of describing what comes back and does so with a concrete field list, which is the key missing piece for an agent. It does not address whether the record can be absent for an unknown id, a minor gap for a 2-param lookup 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 100%, so both parameters (id, lang) are fully documented in the schema, including the id format example and the lang enum's skip behavior. The description reinforces only that the id comes from find_tools and says nothing about lang, so baseline 3 is appropriate.

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+resource ('Full catalog record for one tool id') and explicitly ties the input to a sibling's output ('returned by find_tools'), which cleanly separates it from find_tools itself. An agent can tell instantly that this is the detail-fetch step of a find-then-get flow.

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 'one tool id returned by find_tools' establishes the required precondition and the intended call sequence. It stops short of an explicit 'do not use this to search' exclusion, but the dependency on find_tools output makes the usage context unambiguous.

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 establish readOnlyHint=true, idempotentHint=true and openWorldHint=false, so safety behavior is fully covered structurally. The description adds the returned field set (headline, teaser, date, publisher, source URL), which is useful since no output schema exists, but adds no constraints such as language-skip behavior or rate limits.

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 tight sentence with a parenthetical field list, front-loading the resource and scope with zero filler. Nothing is redundant.

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 usefully enumerates return fields, and the annotations plus full schema coverage handle safety and inputs. It stops short of explaining whether results are ordered by date or how the language-skip behaves, but it is complete enough to invoke correctly.

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%, with enums and defaults documented for lang, limit and category, so the schema carries the parameter burden. The description only nods at the category filter and adds no syntax or semantics beyond what the schema already states – 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+resource ("Newest AI news stories from Base of AGI") and enumerates the returned fields, so the agent knows exactly what it gets. It does not explicitly differentiate itself from the sibling search_news, but "newest" carries an implicit recency scope.

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 only usage cue is "Optional category filter" and the word "Newest", which implies a recency-oriented fetch. There is no guidance on when to prefer this over search_news or daily_brief, so the agent must infer the choice.

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 establish read-only, idempotent, non-destructive behavior, so safety is covered; the description adds real behavioral context on top: archive start (early 2026), the 45-day teaser limitation, AND matching semantics, and ranking (newest first among equal matches). It still omits what a result actually contains and whether results are truncated.

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 tight sentences with zero filler; corpus scope, matching rule, and ordering are each delivered in a single clause, and the most important scoping fact (what is searchable) is front-loaded.

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 four-parameter read tool with no output schema, the description supplies corpus extent, language behavior, match semantics, and ordering — the essentials for calling it correctly. The residual gaps are return-shape hints and explicit routing away from the sibling latest_news.

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 75% schema coverage the baseline is 3, and the description goes beyond it: 'all four languages' clarifies the lang enum's effect, the archive start date frames the days parameter, and the AND rule adds meaning the query schema's example does not convey. Only the undocumented limit parameter is left unaddressed.

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 ('full-text search over the Base of AGI news archive') and bounds the corpus with concrete coverage details. It does not, however, distinguish itself from the sibling latest_news, which is the tool an agent is most likely to confuse it with.

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?

Usage is implied rather than stated: 'All query words must match' tells the agent this is an AND-style keyword search, which is a usable selection cue. There is no explicit when-to-use/when-not, and no pointer to latest_news or daily_brief as the alternative for browsing rather than searching.

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