CtrlK
BlogDocsLog inGet started
Tessl Logo

writing-agent-instructions

How to write great agent instructions for an agent-native app or template: AGENTS.md, skills, and tool/action descriptions. Use when authoring or reviewing AGENTS.md, writing a SKILL.md, wording action descriptions, or deciding what belongs in instructions vs skills vs memory.

62

Quality

75%

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/writing-agent-instructions/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

71%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.

A well-structured, actionable, and largely concise instruction skill with clear sequencing and built-in verification guidance. Its main gap is progressive disclosure: it is a monolithic single-file guide that defers to external skills only by name rather than via its own reference files.

Suggestions

Split exhaustive material (e.g. the full layer-ownership and budget sections) into one-level-deep reference files under references/ and link to them from the SKILL.md overview.

Tighten a few rationale paragraphs that restate the same principle (e.g. the 'say each thing once' and 'prefer judgment' sections overlap) to recover token budget.

Add an explicit validation checkpoint to the workspace-skills sync workflow (run sync then guard) as a numbered validate-then-proceed step.

DimensionReasoningScore

Conciseness

The body is dense and mostly efficient, assuming the reader's competence and avoiding basic concept re-explanation, but a few sections restate principles or pad rationale that could be trimmed.

4 / 5

Actionability

It gives concrete, executable guidance — example AGENTS.md and frontmatter blocks, a layer-ownership table, code snippets — with only minor gaps where advice stays at the principle level.

4 / 5

Workflow Clarity

Sequencing is clear across most workflows (sync steps, the unpack/edit/validate/pack-style budget model), with verification baked in as a core rule, though a couple of multi-step processes lack an explicit checkpoint.

4 / 5

Progressive Disclosure

No bundle files exist, and while the body is well-sectioned and explicitly points to sibling skills (create-skill) for the full format, it is a single long inlined document with no one-level-deep reference files of its own.

3 / 5

Total

15

/

20

Passed

Description

78%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, specific description with an explicit 'Use when' clause that clearly answers both what and when. It is distinguishable from sibling skills and uses natural trigger language, with only minor room to broaden keyword synonyms.

Suggestions

Add a couple of natural synonyms or phrasings users might say (e.g. 'reviewing agent rules', 'improving tool descriptions') to broaden trigger coverage.

Consider naming one or two more concrete actions (e.g. 'structure', 'trim') to push specificity from several actions toward comprehensive coverage.

DimensionReasoningScore

Specificity

Names the domain (writing agent instructions) and several concrete surfaces — 'AGENTS.md, skills, and tool/action descriptions' — with specific authoring tasks, though it stops short of an exhaustive action list.

4 / 5

Completeness

Clearly states what it covers and provides an explicit 'Use when...' clause enumerating four triggering scenarios; the 'when' is concrete but could list a few more natural variations.

4 / 5

Trigger Term Quality

Includes natural trigger phrases users would say ('authoring or reviewing AGENTS.md', 'writing a SKILL.md', 'wording action descriptions'), with good but not exhaustive synonym coverage.

4 / 5

Distinctiveness Conflict Risk

The niche is narrow and clearly bounded to authoring/reviewing agent-instruction surfaces, with triggers unlikely to collide with general coding or unrelated skills.

5 / 5

Total

17

/

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.

Validation13 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

metadata_version

'metadata.version' is missing

Warning

metadata_field

'metadata' should map string keys to string values

Warning

frontmatter_unknown_keys

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

Warning

Total

13

/

16

Passed

Repository
BuilderIO/agent-native
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.