CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

verification-before-completion

Claiming work is complete without verification is dishonesty, not efficiency. Use when ANY variation of success/completion claims, ANY expression of satisfaction, or ANY positive statement about work state.

Quality

72%

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 ./skills/verification-before-completion/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

66%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 skill's core workflow (the Gate Function) is excellent — clearly sequenced with explicit validation and error-recovery branching — and the claim/evidence tables make guidance concrete. However, the body is noticeably padded: the same rule is dramatized and restated across four sections, 'When to Use' appears twice (the second instance is filler), and section content overlaps heavily between Red Flags and the Rationalization table.

Suggestions

Collapse 'The Iron Law', 'The Bottom Line', and the redundant restatements into the Gate Function section so the rule is stated once as the canonical workflow.

Delete the duplicate '## When to Use' section at the end (the filler line 'This skill is applicable to execute the workflow or actions described in the overview') and keep only the substantive earlier version.

Merge the overlapping 'Red Flags - STOP' list and 'Rationalization Prevention' table into a single excuse/red-flag table, and drop or contextualize the 'From 24 failure memories' reference that means nothing to a fresh reader.

DimensionReasoningScore

Conciseness

The core rule is restated at least four times ('The Iron Law' code block, 'The Gate Function', 'Key Patterns', 'The Bottom Line'), a 'When to Use' section appears twice with the second being pure filler ('This skill is applicable to execute the workflow or actions described in the overview'), and 'From 24 failure memories' is contextless padding. It never explains concepts Claude already knows, so it is not severely verbose, but several padded and redundant sections place it noticeably below the midpoint.

2 / 5

Actionability

The Gate Function is a concrete 5-step procedure (IDENTIFY/RUN/READ/VERIFY/ONLY THEN) and the Common Failures table maps each claim to required evidence (e.g., 'Test command output: 0 failures', 'VCS diff shows changes'), with a specific red-green regression recipe. Falls short of fully executable because patterns like '[Run test command] [See: 34/34 pass]' are placeholder sketches rather than literal commands, though the guidance is concrete for an instruction-only skill.

4 / 5

Workflow Clarity

The Gate Function is a clearly sequenced 5-step workflow with an explicit validation checkpoint and error-recovery branching ('VERIFY: Does output confirm the claim? - If NO: State actual status with evidence'), the red-green cycle provides a feedback loop, and the Common Failures table acts as a checklist. Not a destructive or batch operation skill, so no validation cap applies.

5 / 5

Progressive Disclosure

No bundle files exist and the single body is well organized with clear section headers and tables. Minor organization gaps (duplicated 'When to Use' section, ~135 lines of inline content with redundancy that could be trimmed) keep it below the top anchor, but nothing is misfiled or unnavigable.

4 / 5

Total

15

/

20

Passed

Description

66%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 has an explicit and well-structured 'Use when' clause with good natural trigger coverage, and it states its principle clearly. Its weaknesses are a 'what' framed as a slogan rather than the skill's actual behavior, and trigger breadth ('ANY positive statement') that risks firing in nearly every session, blurring distinctiveness.

Suggestions

State what the skill does in concrete terms, e.g., 'Requires fresh verification evidence (command output, exit codes, diffs) before any success or completion claim' instead of only the principle slogan.

Keep the exhaustive 'ANY variation' framing but anchor it with the most common concrete phrasings users and Claude actually produce ('done', 'finished', 'tests pass', 'build works', 'bug fixed') to improve trigger term coverage.

Narrow the trigger scope slightly (e.g., 'before reporting task completion, committing, or expressing that work is done') to reduce overlap with unrelated skills while preserving intent.

DimensionReasoningScore

Specificity

'Claiming work is complete without verification is dishonesty, not efficiency' names the domain clearly, but the description lists no concrete actions the skill performs (run the verification command, read full output, check exit codes). Fits the anchor for a clear domain with limited concreteness; not a 4 because no specific action list is present.

3 / 5

Completeness

Both 'what' (the verification-before-completion-claims principle) and 'when' (an explicit 'Use when...' clause with concrete trigger conditions) are present. The 'what' is stated as a principle rather than what the skill actually does (enforce running verification before claiming status), so it is not fully explicit — matching the anchor where both exist but could be sharper.

4 / 5

Trigger Term Quality

Natural trigger phrases are well covered — 'ANY variation of success/completion claims', 'ANY expression of satisfaction', 'ANY positive statement about work state' — and the 'Use when' framing is explicit. Missing common variations like 'done', 'finished', 'tests pass', or 'fixed', so it falls short of comprehensive synonym coverage.

4 / 5

Distinctiveness Conflict Risk

The niche (verification discipline before completion claims) is distinct from topical skills, but the trigger scope 'ANY positive statement about work state' / 'ANY expression of satisfaction' is broad enough to fire across virtually all work contexts, creating real overlap risk with general workflow and task-management skills. Somewhat specific, but the breadth keeps it from a mostly-distinct rating.

3 / 5

Total

14

/

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

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

Total

15

/

16

Passed

Repository

administrakt0r/AI-Agents-Safe-Coding-Skills

Path

skills/verification-before-completion/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.