CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

ship

Use when the user asks to take a new task all the way: plan, implement, review, commit and push. Not when a ready task file already exists. Self-contained — execute the inlined steps directly, do NOT invoke other skills via the Skill tool.

Quality

78%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

tessl review fix ./plugins/flowai/skills/ship/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

63%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 exceptionally actionable and its phase/gate sequencing is rigorous, but it pays for that with severe verbosity and zero content split out into bundle files, and two copy-paste step-numbering defects blur the sequence. It reads as a generated composite that inlined six skills' worth of defensive rationale without compression.

Suggestions

Split per-phase detail into one-level-deep reference files (e.g. the Review Phase's JiT mutation machinery and the Push Phase's CI-await procedure are self-contained procedures that would live as references), keeping SKILL.md as the gate-and-sequence overview.

Strip dated incident annotations and repeated cross-phase warnings down to a single statement each — rules like the agent-handle redemption and no-`git-stash` guidance appear nearly verbatim in multiple phases.

Fix the malformed step numbering in the Implement Phase ("6. **6. **Hand off...") and Push Phase (step 7's text embedded in step 6's parenthetical and repeated as step 7), which make the intended sequence ambiguous at those points.

DimensionReasoningScore

Conciseness

The ~585-line body is heavily padded with defensive digressions, dated incident annotations ("observed 2026-09-06, `plan-affected-surface-scout`"), and rules repeated across phases (handle-redemption rules appear in Plan step 2 and Review step 3; the no-`git stash` rules in both Implement and Review). It does not teach concepts Claude already knows, which keeps it above the lowest anchor, but the padding goes far beyond "some unnecessary explanation".

2 / 5

Actionability

Fully executable throughout: exact commands (`git status --porcelain`, `git worktree add <scratch-dir> HEAD`), manifest-detection priority lists, exact CI exit-code semantics, verbatim prompt templates, and a concrete verdict-selection order leave no meaningful gaps.

5 / 5

Workflow Clarity

Six phases in strict order with explicit gates, feedback loops (RED→GREEN→REFACTOR→CHECK), and a final verification checklist, but structural copy-paste defects muddy the sequence: Implement Phase's duplicated "6. **6. **Hand off" numbering and Push Phase step 6 inlining step 7's full text inside a parenthetical while step 7 repeats it.

4 / 5

Progressive Disclosure

Headers, gates, and a verification section give the body real structure, but it is a self-contained monolith — no bundle files exist and ~100 KB of material that clearly belongs in separate files (per-phase detail, the JiT mutation machinery, the CI-await procedure) is entirely inlined.

3 / 5

Total

14

/

20

Passed

Description

87%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 strong description: it explicitly states what the skill does, when to use it, and when not to, with a clear composite niche. Its only weaknesses are unmentioned phases (reflect) and missing common synonyms for the trigger concept.

DimensionReasoningScore

Specificity

The description lists several concrete phase actions ("plan, implement, review, commit and push"), but coverage has minor gaps — the reflect phase is never mentioned and the actions stay at phase-level rather than outcome-level.

4 / 5

Completeness

Both what (the five phases) and when are explicitly answered with an explicit "Use when..." clause and a concrete exclusion condition, matching the anchor for clearly answering both with concrete trigger phrases.

5 / 5

Trigger Term Quality

Natural user phrasings like "take a new task all the way" and the phase names are present, plus a negative trigger ("Not when a ready task file already exists"), but common synonyms such as "ship", "end-to-end", or "finish this task" are missing.

4 / 5

Distinctiveness Conflict Risk

A clear niche (full-cycle composite) with distinct triggers and an explicit negative boundary against the atom skills, keeping conflict risk minimal rather than minor.

5 / 5

Total

18

/

20

Passed

Validation

81%

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

Validation — 13 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

skill_md_line_count

SKILL.md is long (586 lines); consider splitting into references/ and linking

Warning

frontmatter_unknown_keys

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

Warning

referenced_paths_exist

Referenced path issues: 4 missing

Warning

Total

13

/

16

Passed

Repository

korchasa/flowai-plugins

Path

plugins/flowai/skills/ship/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.