Tessl • The management layer for AI agent skills

SkillPublic

provider-research

Research every provider behind Pipecat's services for new models and API affordances, writing per-service reports and local branches for clear-cut updates; publishing is scripts/provider-watch/publish.py's job, run outside this skill

Quality

83%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Quality

Content

92%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A highly actionable, well-sequenced orchestration skill with copy-paste commands, explicit failure fallbacks, and clean delegation of researcher detail to one-level-deep reference files. Its only weakness is mild repetition of the publishing boundary and a few trimmable asides.

Suggestions

State the 'publishes nothing' boundary once (e.g., in Guardrails) and reference it tersely elsewhere, saving the repeated full sentences.

Tighten the Step 3 paragraph about Claude Code vs Codex agent definitions into a short fallback note pointing at the two agent-definition paths.

Dimension	Reasoning	Score
Conciseness	Dense procedural instruction with real commands and no concept explanations Claude already knows. The publish boundary ('Everything stays local — this skill publishes nothing' / 'This skill publishes nothing: never push...') is stated three times, and asides like 'the agent definitions are thin shims over that guide' could be trimmed — minor over-explanation, not padded sections.	4 / 5
Actionability	Copy-paste ready commands throughout: 'git rev-parse --short HEAD', 'mktemp -d -t provider-research', the full gh\|jq digest-comments pipeline, 'uv run python scripts/provider-watch/inventory.py --json', an exact JSON payload spec, and fallbacks for common failure cases (clone failure, gh unavailable, researcher failure).	5 / 5
Workflow Clarity	Four clearly sequenced steps with explicit validation checkpoints and error-recovery feedback loops for a batch operation: 'Stop with a clear error if ... inventory.py --md fails', 'If a researcher fails or returns nothing usable, write the report yourself ... never aborts the run', 'If git status ... shows changes you did not make, stop and report it', plus worktree cleanup. No evident validation gaps.	5 / 5
Progressive Disclosure	A clear orchestration overview with well-signaled, one-level-deep references: researcher behavior in 'RESEARCH_GUIDE.md', report format in 'REPORT_TEMPLATE.md', publishing in 'scripts/provider-watch/publish.py', units from inventory.py. Researcher detail is appropriately split out of SKILL.md and the reference discipline is restated as a guardrail.	5 / 5
	Total	19 / 20 Passed

Description

70%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A specific, distinctive description with concrete actions and a clear publish boundary, written in third person. Its main gap is the absence of any explicit 'when to use this' trigger guidance, which caps completeness at 3.

Suggestions

Add a 'Use when...' clause naming the trigger situations (e.g., 'Use when the provider-watch sweep is due, or when you need to check Pipecat providers for new models, pricing, or API changes').

Replace the abstract 'API affordances' with plain capabilities users would name (e.g., 'new models, endpoints, and settings fields').

Mention the digest output or decisions intake so the description covers the full sweep scope, not just reports and branches.

Dimension	Reasoning	Score
Specificity	Lists several concrete actions — 'Research every provider behind Pipecat's services for new models and API affordances', 'writing per-service reports', 'local branches for clear-cut updates' — plus an explicit publish boundary. Falls short of 5 because 'API affordances' is somewhat abstract and the digest/decisions work is unmentioned.	4 / 5
Completeness	The 'what' is clear and concrete, but there is no 'Use when...' clause or equivalent explicit trigger guidance, which caps completeness at 3 per the rubric guideline. Not a 2 because the 'what' is specific rather than vague.	3 / 5
Trigger Term Quality	Good natural keywords for the niche ('Research', 'provider', 'Pipecat's services', 'new models', 'per-service reports', 'local branches'); a few natural terms users might say are missing (e.g., 'LLM', 'STT/TTS', 'speech').	4 / 5
Distinctiveness Conflict Risk	A clear niche ('every provider behind Pipecat's services', 'scripts/provider-watch/publish.py') with distinct triggers and an explicit boundary separating it from publishing skills — minimal conflict risk.	5 / 5
	Total	16 / 20 Passed

Validation

87%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning
referenced_paths_exist	Referenced path issues: 6 missing, 5 deeper-than-1-level	Warning

	Total	14 / 16 Passed

Repository: pipecat-ai/pipecat
Path: .claude/skills/provider-research/SKILL.md
Revision: d49f309

Reviewed: 9 days ago

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.

provider-research

Table of Contents

Is this your skill?