Tessl • The management layer for AI agent skills

SkillPublic

route

The user meets the deficit their context shows as the protocol that resolves it and sees why, without recalling which exist — /route invokes what plain fits compose into, else points or stays silent.

Quality

43%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

tessl review fix ./route/skills/route/SKILL.md

Quality

Content

31%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 presents a genuine, coherently specified one-pass routing procedure, but delivers it as a ~340-line Lean formalization inlined in SKILL.md whose key judgments are axioms, restated across many redundant sections. It functions as a philosophy/contract document rather than actionable skill instructions.

Suggestions

Move the full Lean formalization to references/formalization.md and keep SKILL.md as a concise overview with the decision procedure (read catalog → list fits → lead/raise/fresh/plain/allowed → invoke/point/silence) and the Rules.

Add concrete examples of each outcome: a sample invocation line, a sample pointer line ("↗ /command — reading"), and a sample silence case, so the model can pattern-match real output.

Cut the redundant restatements (MORPHISM, MODE STATE, PHASE TRANSITIONS, and COMPOSITION largely repeat FLOW and the prose Definition) to reduce the body to its essential decision rule and output formats.

Dimension	Reasoning	Score
Conciseness	The same one-pass routing decision is restated repeatedly across the prose Definition, FLOW, MORPHISM, GROUND, TYPES, MODE STATE, PHASE TRANSITIONS, TOOL GROUNDING, COMPOSITION, and Rules sections of a ~360-line body. Several sections are padded restatements, matching the noticeably-verbose anchor rather than the severely-padded one.	2 / 5
Actionability	The Lean block is elaborated code, but every operative judgment (catalog, FitSupported, Plain, lead, Raise, Fresh) is an axiom — a black box — and the Rules give high-level directives ("Route's reading is shown") without concrete executable steps, examples, or templates. It is above level 1 because the decision procedure and output formats are semi-specified, but below level 3 because no copy-usable guidance exists.	2 / 5
Workflow Clarity	FLOW lays out a coherent one-pass sequence (invoke a plain allowed fit / pointer lines / silence) and the five Rules are numbered, so a sequence is present. But it is buried in formalism with no validation checkpoints, feedback loops, or error-recovery guidance, matching the steps-present-but-checkpoints-missing anchor.	3 / 5
Progressive Disclosure	A ~340-line formal Lean specification is inlined monolithically in SKILL.md with zero external reference files (references/, scripts/, assets/ are absent). This is content that clearly belongs in a separate bundled file, matching the minimal-structure anchor; it is not level 1 because the body does have clear section headers.	2 / 5
	Total	9 / 20 Passed

Description

28%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 names a real niche and three outcomes but expresses them in dense private jargon with no "use when" trigger guidance and almost no natural user keywords. It reads as an abstract philosophy statement rather than a capability description a person or model could act on.

Suggestions

Add an explicit trigger clause, e.g. "Use when the user asks for a protocol/skill to be chosen, or a routing decision between loaded skills, or invokes /route."

Replace private jargon ("the deficit their context shows", "what plain fits compose into") with concrete capability verbs and objects, e.g. "Invokes the matching core skill, or lists one-line pointers to candidate skills, or stays silent."

Include natural trigger terms and synonyms users would actually type ("route", "pick a skill", "which protocol", "/route") instead of only the slash-command name.

Dimension	Reasoning	Score
Specificity	The domain (routing to a protocol) is named and three outcome actions appear ("invokes", "points or stays silent"), but they are buried in abstract language like "the deficit their context shows" and "what plain fits compose into", with no concrete graspable capabilities. It is above level 1 because actions are named, but below level 3 because those actions are generic and swamped by jargon.	2 / 5
Completeness	The "what" is present but vague (invoke a fitting protocol, point, or stay silent) and the "when" is entirely absent — there is no "Use when..." clause or equivalent trigger guidance. This matches the vague-what/no-when anchor rather than the clear-what/missing-when anchor at level 3.	2 / 5
Trigger Term Quality	The only trigger term is "/route"; phrases like "deficit", "protocol", and "plain fits compose" are not words a user would naturally say when needing this skill. Not level 1 because a real trigger (/route) exists, but not level 3 because no natural synonyms or common user phrasings appear.	2 / 5
Distinctiveness Conflict Risk	It targets a distinct niche (meta-routing across loaded core protocols), so it is not generic, but the abstract phrasing makes it hard to distinguish from other routing/orchestration skills in practice. It could still overlap with similar skills, matching the level 3 anchor.	3 / 5
	Total	9 / 20 Passed

Validation

100%

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 — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: jongwony/epistemic-protocols
Path: route/skills/route/SKILL.md
Revision: dbf192b

Reviewed: 8 days ago

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.

route

Table of Contents

Is this your skill?