coach

Learning telemetry, strategy, and schedule — retention stats, calibration, grader audit, n-of-1 experiments, HTML dashboard. Use for "how am I doing", weekly check-ins, strategy questions, auditing the grader, or adjusting how Engram teaches.

Quality

88%

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

77%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body is highly actionable and workflows are crisply sequenced with validation checkpoints, but it is a dense monolith with no progressive disclosure into bundle files and noticeable version-history prose that could be trimmed. Splitting the larger procedures into reference files and cutting historical asides would raise the two weaker dimensions.

Suggestions

Move the self-contained `audit`, `experiment`, and `contribute` procedures into one-level-deep reference files (e.g. references/audit.md, references/experiments.md) and keep SKILL.md as an overview with clearly signaled links, to improve progressive_disclosure.

Trim version-history anecdotes and dogfood-discovery asides ("Found by a dogfood", "v0.6 shipped a dead feature that a dogfood certified") into a short changelog or remove them, keeping only the operational rule they motivate, to improve conciseness.

Collapse the repeated emphatic restatements of the same rule (e.g. the multiple "never"/"do not" refrains within a single section) into one stated prohibition to reduce token load without losing the constraint.

Dimension	Reasoning	Score
Conciseness	The 346-line body is mostly purposeful operational detail Claude would not already know, but it carries repeated version-history anecdotes and emphatic rationale (e.g. "v0.6 shipped a dead feature that a dogfood certified", "Found by a dogfood") that pad the token budget beyond what the instruction requires.	2 / 3
Actionability	Every section ships concrete, executable commands (`python3 "$ENGRAM" stats`, `adherence`, `gold > /tmp/engram-gold.json`) plus a complete JSON `experiment start` example, making the guidance copy-paste ready rather than descriptive.	3 / 3
Workflow Clarity	Multi-step procedures are explicitly ordered with validation checkpoints — the audit's three rules, contribute's four non-skippable steps, fail-closed engine resolution, and `doctor` refusing to overwrite a live file — matching the clear-sequence-with-feedback-loops anchor.	3 / 3
Progressive Disclosure	Sections are well-organized with clear headers, but at 346 lines it is a single monolithic file with no bundle files in references/scripts/assets; substantial self-contained procedures (audit, experiment, contribute) are inline that could be split into one-level-deep reference files, fitting the "content that should be separate is inline" anchor.	2 / 3
	Total	10 / 12 Passed

Description

100%

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, third-person, and cleanly answers both what the skill does and when to invoke it, with concrete capability nouns and natural trigger phrases. It is a strong, low-conflict description.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete capabilities — "retention stats, calibration, grader audit, n-of-1 experiments, HTML dashboard" — matching the anchor for enumerating specific concrete actions rather than vague language.	3 / 3
Completeness	Explicitly states both what it does (the telemetry/strategy/schedule capabilities) and when to use it via the "Use for ..." clause, satisfying the what-and-when anchor with explicit triggers.	3 / 3
Trigger Term Quality	Natural user phrases like "how am I doing", "weekly check-ins", "strategy questions", and "auditing the grader" are exactly what a learner would say, giving good coverage of real trigger terms.	3 / 3
Distinctiveness Conflict Risk	The Engram learning-telemetry niche and its grader-audit / n-of-1-experiment triggers are highly specific and unlikely to fire for unrelated skills, fitting the clear-niche anchor.	3 / 3
	Total	12 / 12 Passed

Validation

87%

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 — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning
relative_links	Relative link issues: 1 suspicious	Warning

	Total	14 / 16 Passed

Repository: nagisanzenin/engram
Path: skills/coach/SKILL.md
Commit: c576e9b

Reviewed: 4 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.