CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

update-rtvi

Carry a change to RTVI's models or message handling through the protocol specification, its version, and its tests

Quality

82%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Passed

No findings from the security scan

SKILL.md
Quality
Evals
Security

Quality

Content

88%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.

The content is an exemplary multi-step workflow skill: concrete executable commands, exact file and test identifiers, explicit validation with error-recovery feedback loops, and a checklist. Its only minor weaknesses are a few passages that could be tightened and the absence of any bundle split for the denser step details.

DimensionReasoningScore

Conciseness

The body is dense and procedural throughout — it never explains concepts Claude already knows and mostly assumes competence ("Read the README, then the diff of the branch against main"). Not anchor 5 because a few passages could be trimmed, e.g., the Step 7 paragraph drafting the Protocol Versions table row ("Note that the Note at the top of the RTVI standard page names the current version...") and parts of the Step 8 summary template.

4 / 5

Actionability

Fully executable guidance throughout: copy-paste commands ("git diff main -- src/pipecat/processors/frameworks/rtvi/ src/pipecat/frames/", the four "uv run" commands in Step 6, the mktemp/worktree sequence in Step 7), exact identifiers ("min_client_version=\"<new version>\"", "self._client_is_at_least([X, Y, 0])"), and precise test and file names ("test_every_behavior_has_a_test", "scripts/deprecations/deprecations.json"). Specific examples cover the common cases for each change kind.

5 / 5

Workflow Clarity

A clear 8-step sequence with explicit validation in Step 6 (pytest, ruff, pyright, and "Check that git status shows no change to a released version's folder"), explicit feedback loops for error recovery ("When a test fails, find what it means before changing anything"; "A released version's contract hash changed: ... Go back to Step 2"), stop-and-ask checkpoints in Step 2, and a closing Checklist plus Guardrails. This matches the anchor-5 example's validate -> fix -> retry pattern exactly.

5 / 5

Progressive Disclosure

No bundle files exist (references/, scripts/, assets/ are absent), so this scores structure alone: clear sections (Arguments, Instructions Steps 1-8, Guardrails, Checklist) and one well-signaled, one-level-deep external reference — "spec/README.md is the source of truth ... Read it first" with explicit non-restatement ("This skill applies the README to one change; it doesn't restate it") and delegation to sibling skills (changelog, update-docs). Not anchor 5 because the ~165-line file keeps all per-step detail inline (version decision rules, docs preview procedure, summary template); not anchor 3 because the structure and signaling are genuinely good.

4 / 5

Total

18

/

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.

The description is specific and unambiguous about its niche — carrying RTVI model/message-handling changes through the spec, version, and tests — with strong distinctiveness. Its main weakness is the missing explicit 'Use when...' trigger guidance, which caps completeness and leaves activation conditions only implied.

Suggestions

Add an explicit trigger clause, e.g., "Use when a change touches models.py or what RTVIObserver sends or RTVIProcessor accepts" — without it, completeness is capped at 3.

Mention the remaining actions the skill performs (gating new output for older clients, regenerating the document, previewing the docs) so the capability list is comprehensive rather than partial.

Include natural trigger phrases users would actually say, such as "RTVI protocol version", "spec", or "conformance tests", to broaden keyword coverage with synonyms.

DimensionReasoningScore

Specificity

Names the domain ("RTVI's models or message handling") and three concrete targets ("the protocol specification, its version, and its tests"), which goes beyond the 1-2 actions of anchor 3. Not anchor 5 because coverage has gaps: gating output for older clients, regenerating the document, and previewing the docs are all performed by the skill but unmentioned.

4 / 5

Completeness

Has a clear 'what' ("Carry a change to RTVI's models or message handling through the protocol specification, its version, and its tests") but no 'Use when...' clause or equivalent; the 'when' is only weakly implied by 'a change to RTVI's models or message handling'. Per the judging guidelines, a missing explicit trigger clause caps completeness at 3.

3 / 5

Trigger Term Quality

Good keyword coverage with natural domain terms a developer would say: "RTVI", "models", "message handling", "protocol specification", "version", "tests". Not anchor 5 because no synonyms or variations appear (e.g., "spec", "conformance", "wire format", "AsyncAPI").

4 / 5

Distinctiveness Conflict Risk

"RTVI" is a highly specific niche term with distinct triggers; the description could not plausibly fire for another skill. This matches anchor 5 (clear niche, minimal conflict risk) rather than anchor 4, which requires some overlap risk with closely related skills.

5 / 5

Total

16

/

20

Passed

Validation

93%

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

Validation — 15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

referenced_paths_exist

Referenced path issues: 2 possibly missing (scripts/deprecations/deprecations.json, scripts/rtvi/write_rtvi_spec.py), 2 deeper-than-1-level

Warning

Total

15

/

16

Passed

Repository

pipecat-ai/pipecat

Path

.claude/skills/update-rtvi/SKILL.md

Reviewed

Table of Contents

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.