CtrlK
BlogDocsLog inGet started
Tessl Logo

thinking-out-loud

A contract for what the agent does when a long, messy, stream-of-consciousness ramble arrives (usually voice dictation): act on nothing until the echo brief is approved. The echo audits the entire transfer, mission, locked decisions and constraints, open questions, flips and parked tangents, with the model's inferences and guesses quarantined away from the user's own phrasing, so the user verifies what the model believes, not just what it doubts. Use when the user says "let me think out loud" or wants to ramble a bit before building anything, when a message opens with a speech-to-text preamble like "switching to voice, sorry for typos", when input is a long weakly punctuated stream with restarts and mid-message reversals ("actually no, wait, scrap that idea entirely"), or when the user asks to be interviewed about a fuzzy half-formed idea. Includes an optional capture mode for rambles spread across several messages and an optional targeted interview.

75

Quality

92%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Critical

Do not install without reviewing

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.

A well-engineered behavioral contract: every rule is imperative and testable, the workflow is gated by explicit user approval with feedback loops in the interview and persistence stages, and the single reference file is real, purposeful, and one level deep. The two weaknesses are mild: ~25 lines of design rationale in the opening sections, and compression-rule duplication between the body and the reference file.

DimensionReasoningScore

Conciseness

The body is overwhelmingly imperative and dense ('Label every addition', 'One line per bullet', 'Exit on "done", "echo", "echo me"'), and it teaches a novel contract rather than concepts Claude already knows. What keeps it below 5 is motivational prose that could be trimmed: the opening two paragraphs ('A ten minute voice ramble transfers more context than any prompt a person would type...') and the 'Why an echo instead of follow-up questions' section argue for the design rather than instruct, roughly 25 lines of rationale. It is well above the 3-anchor's 'unnecessary explanation' since none of it re-teaches known concepts.

4 / 5

Actionability

For an instruction-only skill the guidance is fully executable: the five-section echo structure with per-section semantics, non-negotiable compression rules, exact acknowledgment lines ('Go ahead, I'm listening. Say "done" when you want the echo.'), enumerated exit phrases, a one-question-per-message interview protocol capped at five, and three concrete persistence options including an exact path format ('docs/rambles/YYYY-MM-DD-<topic>.md'). The worked example in the reference makes the common case copy-paste ready.

5 / 5

Workflow Clarity

The sequence is explicit and gated: reconstruct first, 'Act on nothing... until the echo is approved' (an explicit validation checkpoint on the risky action of building from a misread), then optional interview with a stop-early condition ('stop early once answers stop changing the brief') and a restatement loop ('restate only the sections of the echo that changed'), then a persistence decision. Capture mode adds its own entry, mid-stream, and exit rules. These are the feedback loops and checkpoints the 5-anchor asks for, with no destructive-operation validation gap.

5 / 5

Progressive Disclosure

Structure is good: clearly-headed sections, and the one external reference is real, one level deep, purpose-stated, and holds exactly the right material ('Full template with a worked example in [references/echo-format.md]'). What keeps it at 4 rather than 5 is that the body is more than an overview — the four compression rules are stated in full in both SKILL.md and echo-format.md, a minor duplication that could be consolidated — and the body at ~135 lines carries contract, interview, capture, and persistence detail that partially belongs behind the reference.

4 / 5

Total

18

/

20

Passed

Description

96%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 dense, trigger-rich description that explicitly answers what the skill does and when to use it, with verbatim user phrasings as triggers. The only deductions are minor: the interview trigger overlaps slightly with general ideation skills, and the description is long, though every clause carries information rather than padding.

DimensionReasoningScore

Specificity

The description lists multiple concrete actions: 'produce an echo', 'audits the entire transfer, mission, locked decisions and constraints, open questions, flips and parked tangents', 'inferences and guesses quarantined away from the user's own phrasing', plus 'an optional capture mode' and 'an optional targeted interview'. Coverage within its niche is comprehensive and every action is stated concretely rather than abstractly.

5 / 5

Completeness

Both halves are explicit: the 'what' ('act on nothing until the echo brief is approved', with a full inventory of what the echo audits) and the 'when' ('Use when the user says "let me think out loud"...', voice preamble, long weakly punctuated stream with restarts). This matches the 5-anchor example structure of capability statement followed by concrete 'Use when' triggers.

5 / 5

Trigger Term Quality

Triggers are quoted verbatim natural user phrasings: '"let me think out loud"', 'wants to ramble a bit before building anything', 'a speech-to-text preamble like "switching to voice, sorry for typos"', '"actually no, wait, scrap that idea entirely"', 'asks to be interviewed about a fuzzy half-formed idea'. These are exactly the phrases a rambling or dictating user would produce, including the mid-message reversal quote that is the most natural signal of this skill's niche.

5 / 5

Distinctiveness Conflict Risk

The ramble/dictation/voice-preamble niche is distinctive and the quoted triggers would not fire for ordinary requests, but 'asks to be interviewed about a fuzzy half-formed idea' has minor overlap risk with general brainstorming or clarification skills. This is a clear step above the 3-anchor ('could still overlap with similar skills') and just below the 5-anchor's 'minimal conflict risk'.

4 / 5

Total

19

/

20

Passed

Validation

100%

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

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository
Shubhamsaboo/awesome-llm-apps
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.