CtrlK
BlogDocsLog inGet started
Tessl Logo

context-fundamentals

Context is the complete state available to a language model at inference time. It includes everything the model can attend to when generating responses: system instructions, tool definitions, retrieved documents, message history, and tool outputs.

32

Quality

28%

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 ./plugins/AI-Agents-Safe-Coding-Skills/skills/context-fundamentals/SKILL.md

The canonical home for this skill is context-fundamentals in administrakt0r/AI-Agents-Safe-Coding-Skills

SKILL.md
Quality
Evals
Security

Quality

Content

38%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 a well-organized but verbose conceptual primer that explains fundamentals Claude largely already knows, with limited executable guidance and a dangling reference to a file that does not exist. It reads as reference material more than an actionable skill.

Suggestions

Cut the definitional teaching of concepts Claude already knows (what a system prompt is, attention n² mechanics, position encoding) and keep only the non-obvious framing and concrete rules.

Either create the referenced 'Context Components Reference' file or remove the dangling reference; turn vague external-resource bullets into actual links or delete them.

Add a concrete, copy-paste-ready procedure (e.g. a context-budget audit checklist with specific commands) to lift actionability and workflow clarity.

DimensionReasoningScore

Conciseness

The ~180-line body extensively teaches concepts Claude already knows (what system prompts are, attention n² relationships, position encoding), with only a few genuinely additive framing bits (the 83.9% observation stat, the altitude balance, 70-80% compaction), placing it between the 'severely verbose/padded' anchor 1 and the 'noticeably verbose' anchor 2.

2 / 5

Actionability

It offers some concrete rules ('Implement compaction triggers at 70-80% utilization', 'Place critical information at attention-favored positions') and template examples, but no executable commands or copy-paste-ready instructions, fitting the 'some concrete guidance but incomplete' anchor rather than the executable anchor 4.

3 / 5

Workflow Clarity

As a conceptual/reference skill with no destructive or batch operation, no validation cap applies; the content is organized into a clear topical progression (anatomy → mechanics → quality → practical → guidelines) but contains no real sequenced workflow with checkpoints, matching the 'steps/sequence present but checkpoints missing' anchor.

3 / 5

Progressive Disclosure

No bundle files exist in references/scripts/assets, yet the body cites a non-existent 'Context Components Reference' and lists vague unlinked external resources, while detailed reference material (anatomy, mechanics) is inlined rather than split into files — matching the 'minimal structure; content that belongs in separate files is inlined' anchor.

2 / 5

Total

10

/

20

Passed

Description

17%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 reads as a dictionary definition of context rather than a skill capability statement: it enumerates what context is but states no actions and gives no trigger guidance. It is unlikely to be surfaced for the right moment and overlaps heavily with related skills.

Suggestions

Rewrite the description around verbs the skill performs (e.g. 'Designs, audits, and optimizes agent context…') instead of defining what context is.

Add an explicit 'Use when…' clause naming concrete user moments (e.g. 'Use when designing a new agent, debugging context-related failures, or reducing token cost').

Include natural trigger terms users would say ('context window', 'token budget', 'prompt is too long', 'agent keeps forgetting') rather than abstract jargon.

DimensionReasoningScore

Specificity

The description richly names the domain and enumerates concrete components ('system instructions, tool definitions, retrieved documents, message history, and tool outputs'), but these are parts of a concept rather than actions the skill performs — it never uses a single capability verb, fitting between the domain-naming anchor 2 and the 1-2-actions anchor 3.

2 / 5

Completeness

It provides a vague 'what' (a definition of context rather than a skill capability) and entirely lacks any 'when'/Use-when guidance, matching the 'vague what and no when' anchor rather than the clear-what anchor 3.

2 / 5

Trigger Term Quality

All keywords are abstract technical jargon ('inference time', 'tool definitions', 'retrieved documents', 'message history', 'tool outputs') with no natural phrases a user would actually say when they need this skill, matching the 'no natural keywords; only technical jargon' anchor.

1 / 5

Distinctiveness Conflict Risk

'Context' as a framing is extremely broad with high overlap risk against the sibling context-engineering skills it itself lists (context-degradation, context-optimization, multi-agent-patterns, tool-design), matching the 'very broad; high overlap risk' anchor.

2 / 5

Total

7

/

20

Passed

Validation

93%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

frontmatter_unknown_keys

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

Warning

Total

15

/

16

Passed

Repository
administrakt0r/AI-Agents-Safe-Coding-Skills
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.