CtrlK
BlogDocsLog inGet started
Tessl Logo

context-compression

When agent sessions generate millions of tokens of conversation history, compression becomes mandatory. The naive approach is aggressive compression to minimize tokens per request.

48

Quality

51%

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-compression/SKILL.md

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

SKILL.md
Quality
Evals
Security

Quality

Content

57%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.

This is a well-structured conceptual skill with concrete templates and steps, but it is verbose with repeated material, lacks validation checkpoints in its workflows, and keeps all content in a single file with a dangling internal reference. Tightening and splitting would lift it substantially.

Suggestions

Dedupe the repeated structured-summary/tokens-per-task material across Detailed Topics, Practical Guidance, Guidelines, and Integration to reduce token cost.

Add explicit validation checkpoints to the workflows (e.g., 'verify artifact trail completeness after each compression cycle') to satisfy the destructive-operation feedback-loop expectation.

Move the evaluation framework detail into an actual references/ file and link it with a real path, or remove the dangling 'Internal reference' line.

DimensionReasoningScore

Conciseness

The body carries domain-specific value but is noticeably verbose: the Detailed Topics, Practical Guidance, Examples, Guidelines, and Integration sections restate the same points (e.g., structured summaries and tokens-per-task), so it fits 'mostly efficient but could be tightened.'

3 / 5

Actionability

Though code-free, it provides copy-paste-ready structured-summary templates, a 5-step implementation recipe, and concrete trigger/evaluation tables — actionable guidance with only minor gaps, landing at 4 rather than 5.

4 / 5

Workflow Clarity

The three-phase workflow and 5-step recipe are sequenced, but neither embeds validation/verification checkpoints; because compression is information-lossy the rubric's destructive-operation cap holds it at 3.

3 / 5

Progressive Disclosure

Everything lives in one ~260-line monolithic file, and the 'Evaluation Framework Reference' is named but points to no real bundle file, matching 'some structure but references not clearly signaled and content that should be separate is inline.'

3 / 5

Total

13

/

20

Passed

Description

45%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 frames a problem scenario rather than stating what the skill does or when to invoke it, leaving it without an explicit 'Use when' trigger. It is domain-specific enough to be distinguishable but lacks concrete capability actions and natural trigger phrasing.

Suggestions

Rewrite to lead with what the skill does ('Compresses/summarizes agent conversation history…') instead of only stating the problem.

Add an explicit 'Use when…' clause with natural triggers (e.g., 'when sessions hit context limits', 'when an agent forgets which files it modified').

List 2-3 concrete capabilities (e.g., structured summary templates, probe-based evaluation) to raise specificity beyond naming the domain.

DimensionReasoningScore

Specificity

The description names the domain ('compression' of 'conversation history') but states no concrete actions the skill performs, only framing a problem and a naive approach; it does not reach the '1-2 concrete actions' bar of a 3.

2 / 5

Completeness

It gives a scenario-style 'when' and a weakly implied 'what' but no explicit trigger guidance ('Use when…'), so per the rubric cap it stays at 3 rather than 4.

3 / 5

Trigger Term Quality

It includes relevant natural terms ('agent sessions', 'conversation history', 'compression', 'tokens') but misses common variations/synonyms and offers no 'Use when…' phrasing, fitting 'some relevant keywords but missing common variations.'

3 / 5

Distinctiveness Conflict Risk

The context-compression niche is somewhat specific, but the generic framing could overlap with closely related skills like context-optimization, matching 'somewhat specific but could still overlap.'

3 / 5

Total

11

/

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.