CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

wf-plan

This skill should only be used when the user uses the word workflow and asks to create a commit-ready, deep project plan from a shaped packet (brief, breadboard, risks, spikes) before development starts, with handoff/pickup boundaries to avoid context rot.

Quality

78%

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 ./.agents/skills/03-plan/wf-plan/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

78%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 efficient, clearly sequenced planning workflow with explicit gates and go/no-go criteria, which is impressive for an instruction-only skill. Its main structural flaw is that the two bundle reference files (plan template, review checklist) are completely unlinked, so the model must rediscover them and the inline summaries duplicate bundle content instead of pointing to it.

Suggestions

Link the bundle files where their content is summarized: in step 2, "Create plan.md following references/plan-template.md" and in step 6, "Run each pass using references/plan-review-checklist.md".

Add a brief note on what invoking `pickup`, `oracle`, and `handoff` should produce (or where their definitions live) so the mandatory passes are self-contained.

State the failure path for the commit gate, e.g. "If any GO condition fails, return to the relevant step (5 or 6) and iterate — do not hand off on a NO-GO."

DimensionReasoningScore

Conciseness

The body is lean and directive throughout — terse bullet instructions like "Read brief → breadboard → risk register → spikes", "Record concrete file paths", and a five-line commit gate with no concept explanations Claude already knows and no padding. Every line instructs; it matches the 'every token earns its place' anchor and is clearly above anchor 4, where minor over-explanation would remain.

5 / 5

Actionability

Mostly executable guidance for an instruction-only skill: named input files (brief.md, breadboard-pack.md, risk-register.md), an enumerated plan.md skeleton, explicit external-research conditions ("security/auth, payments, privacy, external APIs, migrations"), a five-condition GO gate, and a handoff content list. It falls short of anchor 5 because key mechanics are delegated to undefined invocations ("invoke `pickup`", "invoke `oracle`", "invoke `handoff`") with no pointers, and review passes like "Simplicity, risk, ops/release" give no pass/fail criteria in the body; it is above anchor 3 because the guidance is concrete and specific rather than pseudocode-level.

4 / 5

Workflow Clarity

A clearly numbered 0-8 sequence with routing on bad inputs ("If perimeter or top risks are missing: route back to wf-shape"), conditional branches, a commit gate, a Verification section, and an explicit Go/No-Go with a concrete NO-GO condition. It is below anchor 5 because the review passes in step 6 lack stated validation criteria in the body (the checklist exists only as an unreferenced file) and no explicit fix-and-retry loop is described for a failed commit gate; it is above anchor 3 because checkpoints (commit gate, GO/NO-GO, oracle pass) are explicit rather than implicit.

4 / 5

Progressive Disclosure

The body is well-sectioned (Purpose, Inputs, Outputs, Steps, Verification, Go/No-Go) but never links to either bundle file: references/plan-template.md and references/plan-review-checklist.md are not mentioned anywhere, while their content is re-summarized inline (the plan.md skeleton in step 2 and the review dimensions in step 6). This matches the anchor 'references present but not clearly signaled; content that should be separate is inline' — it is above anchor 2 because the inlined content is a compressed overview rather than the full files dumped into the body, and below anchor 4 because not a single reference path is surfaced.

3 / 5

Total

16

/

20

Passed

Description

67%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 explicitly covers both what the skill does and when to use it, with concrete artifact-level specificity and a well-bounded trigger. Its main weaknesses are heavy reliance on internal jargon (shaped packet, breadboard, context rot) that users would not naturally say and a single-word trigger that misses common phrasings like implementation plan or tech spec.

Suggestions

Add natural user phrasings as trigger synonyms, e.g. "Use when the user says workflow and asks for a project plan, implementation plan, or build plan."

Reduce internal jargon in the trigger path (e.g. explain 'shaped packet' as 'a completed brief/breadboard/risk-register packet' or move terms like 'context rot' out of the when-clause).

State the output more concretely as an action pair, e.g. "Reads the packet, researches the repo, and writes a commit-ready plan.md with acceptance criteria, phases, and rollout/rollback."

DimensionReasoningScore

Specificity

The description lists several concrete elements — "create a commit-ready, deep project plan from a shaped packet (brief, breadboard, risks, spikes)" and "handoff/pickup boundaries" — naming specific input artifacts and output, with only minor coverage gaps (e.g., what the plan contains). It falls below anchor 5 because it essentially describes one action (create a plan) rather than multiple comprehensive concrete actions, and above anchor 3 because the artifact-level detail goes beyond naming the domain with 1-2 actions.

4 / 5

Completeness

Both are explicitly present: the what ("create a commit-ready, deep project plan from a shaped packet... with handoff/pickup boundaries") and the when ("should only be used when the user uses the word workflow and asks to create a commit-ready, deep project plan"). It is below anchor 5 because the when-clause is a single narrow word-trigger wrapped in conditional meta-language ("This skill should only be used when...") rather than clean, varied concrete trigger phrases, and above anchor 3 because the when is fully explicit, not weakly implied.

4 / 5

Trigger Term Quality

"workflow" and "plan / project plan" are natural terms a user would say, but much of the description is internal system jargon users would not naturally utter ("shaped packet", "breadboard", "spikes", "handoff/pickup boundaries", "context rot"), and common variations like "implementation plan", "tech spec", or "planning" are missing. It is not anchor 2 because the explicit word-trigger "workflow" plus "asks to create a... plan" are genuinely relevant keywords, not generic filler; it is not anchor 4 because the natural-term coverage is thin relative to the jargon load.

3 / 5

Distinctiveness Conflict Risk

The restriction to "when the user uses the word workflow" combined with the specific upstream inputs ("shaped packet (brief, breadboard, risks, spikes)") carves out a distinct niche that is unlikely to fire for unrelated skills. It is not anchor 5 because "workflow" + "plan" still overlaps with generic planning/spec-writing skills and sibling workflow-family skills, leaving minor overlap risk.

4 / 5

Total

15

/

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

marchatton/agent-skills

Path

.agents/skills/03-plan/wf-plan/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.