system-prompts

Write system prompts, tool docs, and agent definitions. Project tag conventions + RFC 2119 keywords + dense compression. Use when authoring or editing any prompt the model reads.

Quality

85%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

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 exemplifies its own house style — dense, table-driven, and immediately actionable with concrete rewrite pairs, explicit placement sequences, and a verification checklist. Its main structural weakness is progressive disclosure: the skill is a single monolithic file whose one MUST-read reference (small-models.md) does not exist in the bundle, and large sections could be offloaded to reference files.

Suggestions

Add small-models.md to the bundle (e.g. under references/) or remove the "MUST read" directive — the link is currently dangling and instructs agents to read a file that does not exist.

Split the ~40-line 'Tool Prompt Authoring' section into a separate reference file (e.g. references/tool-prompts.md) with a one-line pointer from SKILL.md, keeping the overview lean.

Add an explicit feedback loop to the Checklist (e.g. 'Re-check each item after edits; fix violations and re-verify') to close the workflow_clarity gap.

Dimension	Reasoning	Score
Conciseness	The body is aggressively dense: table-driven tag and RFC-keyword references, compressed bullets ("Missing? Re-`read`."), and paired Bad/Good rewrites with zero padding. It assumes Claude's competence (no explanation of what a system prompt is or how models work) and every line shifts a decision, matching the 5 anchor 'every token earns its place'.	5 / 5
Actionability	Guidance is copy-paste ready: the RFC table includes a 'Replaces' column mapping hedges to keywords, front-matter/back-matter give numbered placement sequences, tool-prompt anatomy is a numbered 6-step structure, and the Bad/Good pairs are concrete rewrite examples an agent can apply directly. For an instruction-only skill this fully covers the common cases, matching the 5 anchor.	5 / 5
Workflow Clarity	Sequences are explicit (front matter 1–5, back matter 1–3, tool-prompt anatomy 1–6) and the final Checklist provides a verification path ("Name verification path (tests, lint, typecheck); NEVER 'review your work'"). It stops short of the 5 anchor because there is no explicit feedback loop — no 'run the checklist, fix violations, re-check' iteration — leaving most-but-not-all checkpoints.	4 / 5
Progressive Disclosure	Sections are clearly headed and well-organized, but the body is a ~180-line monolith: the body's only external reference — "Small models (≤2B; tiny/on-device, e.g. LFM2): MUST read [small-models.md](small-models.md)" — is a dangling link, with no small-models.md in the bundle (no references/, scripts/, or assets/ directories exist). Substantial material that would naturally split out (e.g. the ~40-line Tool Prompt Authoring section) is inlined, fitting the 3 anchor ('content that should be separate is inline') rather than the 4 anchor's 'minor organization gaps'.	3 / 5
	Total	17 / 20 Passed

Description

83%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 is concise, concrete, and complete — it states what artifacts the skill produces, what methodology it enforces, and an explicit 'Use when' trigger clause. Keyword coverage is good, though it omits a few natural synonyms such as 'prompt engineering', and the broad 'any prompt the model reads' trigger creates minor overlap risk with general writing skills.

Suggestions

Add common natural synonyms to the trigger clause, e.g. 'Use when authoring or editing system prompts, tool documentation, agent definitions, or doing prompt engineering work'.

Tighten the trigger scope from 'any prompt the model reads' to the specific artifact types already named, reducing overlap with general writing/editing skills.

Dimension	Reasoning	Score
Specificity	"Write system prompts, tool docs, and agent definitions" names the domain and three concrete deliverables, with the methodology compressed into "Project tag conventions + RFC 2119 keywords + dense compression". It falls just short of the 5 anchor because the actions are really one verb (write/author) applied to artifact types rather than multiple distinct comprehensive actions.	4 / 5
Completeness	Explicitly answers what ("Write system prompts, tool docs, and agent definitions. Project tag conventions + RFC 2119 keywords + dense compression") and when with a concrete trigger phrase ("Use when authoring or editing any prompt the model reads"). The 'when' is explicit and specific, so it matches the 5 anchor rather than the 4 anchor's 'when could be more explicit or specific'.	5 / 5
Trigger Term Quality	Includes natural terms a user would say — "system prompts", "tool docs", "agent definitions", "authoring", "editing", "prompt". A few common variations are missing (e.g. "prompt engineering", "instructions", "persona"), so it fits the 4 anchor (good coverage, a few natural terms missing) rather than the 5 anchor's comprehensive synonym coverage.	4 / 5
Distinctiveness Conflict Risk	A clear niche (authoring prompts the model reads) with distinct triggers around system prompts, tool docs, and agent definitions. The broad phrase "any prompt the model reads" leaves minor overlap risk with general writing/editing skills, placing it at 4 rather than the 5 anchor's minimal conflict risk.	4 / 5
	Total	17 / 20 Passed

Validation

93%

Warnings & errors only

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

Criteria	Description	Result
relative_links	Relative link issues: 1 missing	Warning

	Total	15 / 16 Passed

Repository: can1357/oh-my-pi
Path: .omp/skills/system-prompts/SKILL.md
Commit: cabfa74

Reviewed: 3 days ago

Table of Contents

Discovery Implementation Validation

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.