CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

proposer-guide

Use source and execution tools to investigate a supplied program and check a proposed change against evidence.

Quality

74%

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

Fix and improve this skill with Tessl

tessl review fix ./agentpilot/src/agentpilot/policy/proposer/skills/proposer-guide/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

86%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 body is an exemplary lean instruction-only skill: concrete tool signatures, explicit epistemic checkpoints for testing a proposed change, and a clear revise-on-counterevidence feedback loop. The only meaningful gap is the absence of a worked example showing the full investigation sequence end to end.

DimensionReasoningScore

Conciseness

The ~13-line body is lean with no padding and no explanation of concepts Claude already knows — every line ('Executor explanations and printed self-verdicts are claims', 'keep an unknown origin unknown') carries non-obvious, domain-specific guidance. Matches 'lean and efficient; assumes Claude's competence; every token earns its place'; not 4 because there is no over-explanation to trim.

5 / 5

Actionability

Concrete, parameterized tool invocations are given ('read(path="src/main.py", line=40)', 'search(query="text")', 'check_trace(ref, start, end, name)') plus explicit behavioral directives like 'check whether the change already exists, would trigger, or would still accept the recorded failure'. Matches 'mostly executable guidance with minor gaps'; not 5 because there is no worked end-to-end example and the evaluation heuristics are stated abstractly without illustration.

4 / 5

Workflow Clarity

A clear implicit sequence runs read missing code → gather tool evidence → inspect attempts → test provisional ideas → 'revise the mechanism or narrow its applicability' on counterevidence, with explicit checkpoints ('Check whether its new action would occur in time to affect that failure'). Matches 'clear sequence with most checkpoints present'; not 5 because steps are unnumbered and error-recovery guidance beyond 'revise' is thin; not 3 because the checkpoints are explicitly articulated rather than implicit.

4 / 5

Progressive Disclosure

The skill is under 50 lines, has no bundle files (no references/, scripts/, or assets/ directories exist), and needs no external references; the body is organized with a header, a tool-usage list, and principle prose. Per the rubric's simple-skill note, this earns 5 with well-organized sections; not 4 because nothing that belongs in a separate file is inlined.

5 / 5

Total

18

/

20

Passed

Description

50%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 communicates a clear, concrete purpose but omits any 'when to use' trigger guidance and leans on internal jargon ('execution tools', 'evidence') rather than natural user phrasing. It is distinguishable in intent yet carries moderate overlap risk with generic investigation/review skills.

Suggestions

Add an explicit trigger clause, e.g. 'Use when evaluating a proposed change or fix to a supplied program, or when verifying claims about how the code actually behaved.'

Replace jargon-heavy phrasing ('source and execution tools') with natural keywords users would say, such as 'read the source code', 'inspect execution traces', 'verify a proposed change'.

Sharpen the capability list with one or two more concrete actions (e.g. 'search source and recorded executions, compare successful and failed attempts') to reduce overlap with generic debugging/review skills.

DimensionReasoningScore

Specificity

Names the domain ('investigate a supplied program') and two actions ('investigate', 'check a proposed change against evidence'), but the tooling is generic ('source and execution tools') and coverage is not comprehensive — matching 'names domain and 1-2 concrete actions'. Not 4 because it does not list several distinct specific actions; not 2 because the actions are concrete rather than minimal or generic.

3 / 5

Completeness

The 'what' is clear ('Use source and execution tools to investigate a supplied program and check a proposed change against evidence') but there is no 'Use when...' clause or equivalent trigger guidance, which caps completeness at 3 per the judging guidelines. Not 4 because the 'when' is entirely missing rather than merely imprecise.

3 / 5

Trigger Term Quality

Some relevant natural phrases appear ('proposed change', 'investigate', 'source'), but jargon like 'execution tools' and 'evidence' dominates and common user variations (e.g. 'review a fix', 'verify a change', 'debug the program') are absent. Fits 'some relevant keywords but missing common variations or synonyms' rather than the good coverage of level 4.

3 / 5

Distinctiveness Conflict Risk

The niche of verifying a proposed change against recorded execution evidence is fairly specific, but 'investigate a supplied program' overlaps with debugging and code-review skills and no distinct triggers are given. Fits 'somewhat specific but could still overlap with similar skills'; not 4 because the overlap risk with review/debugging skills is more than minor.

3 / 5

Total

12

/

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

HKUDS/OpenSpace

Path

agentpilot/src/agentpilot/policy/proposer/skills/proposer-guide/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.