common-context-optimization

Maximize context window efficiency, reduce latency, and prevent lost-in-middle issues through strategic masking and compaction. Use when token budgets are tight, tool outputs overflow the context, conversations drift from intent, or latency spikes from cache misses.

Quality

76%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Critical

Do not install without reviewing

Fix and improve this skill with Tessl

tessl review fix ./.github/skills/common/common-context-optimization/SKILL.md

Quality

Content

68%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 concise, well-structured overview that defers detail to real reference files and gives concrete thresholds. Its main weakness is workflow clarity: destructive masking and compaction operations proceed without validation or verification checkpoints.

Suggestions

Add validation/verification steps to the destructive workflows: after masking, confirm no required data was lost before discarding the raw output; after compaction, verify the preserved state still captures the active goal and current errors.

Add a brief recovery/rollback note for compaction (e.g., keep the pre-compaction turn accessible until the next user turn confirms intent is preserved).

De-duplicate reference pointers: link each reference once in its section and consolidate the rest in the References list to tighten navigation.

Dimension	Reasoning	Score
Conciseness	The body is lean with a tight Problem/Solution/steps format and no padding about concepts Claude already knows; minor redundancy comes from duplicate 'See references/...' lines within sections that also appear in the References list.	4 / 5
Actionability	Concrete thresholds (50 lines or 1 KB, every 10 turns or 8k tokens, System -> Tools -> RAG -> User) and specific steps make guidance mostly executable, though the actual masking mechanism ('rewriting history') is deferred to references rather than shown inline.	4 / 5
Workflow Clarity	Observation Masking and Context Compaction present numbered sequences, but both are destructive operations (replacing raw output, dropping dialogue) with no validation or verification checkpoints, capping this dimension at 3 per the rubric.	3 / 5
Progressive Disclosure	The body is an overview with well-signaled, one-level-deep, real references (masking.md, compaction.md, implementation.md) and a dedicated References section; minor organization gaps from implementation.md being linked ad hoc within multiple sections.	4 / 5
	Total	15 / 20 Passed

Description

83%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 well-constructed description that clearly states capabilities and explicit 'Use when' triggers with concrete, natural phrases. It is comprehensive on completeness and distinctiveness, with only minor gaps in action specificity and trigger synonym coverage.

Suggestions

Tighten the 'what' clause toward concrete outcomes (e.g., 'mask large tool outputs into summaries and compact long conversations to preserved state') so capabilities read as deliverables rather than techniques.

Add a couple of natural user synonyms to the trigger clause (e.g., 'summarize chat history', 'free up context') to broaden keyword coverage.

Dimension	Reasoning	Score
Specificity	Names several concrete actions ('strategic masking and compaction', 'reduce latency', 'prevent lost-in-middle issues') but the actions are technique-level rather than fully concrete outcomes, leaving minor coverage gaps.	4 / 5
Completeness	It explicitly answers both what ('Maximize context window efficiency, reduce latency, and prevent lost-in-middle issues through strategic masking and compaction') and when ('Use when token budgets are tight...') with concrete trigger phrases.	5 / 5
Trigger Term Quality	The 'Use when token budgets are tight, tool outputs overflow the context, conversations drift from intent, or latency spikes from cache misses' clause plus metadata keywords give good natural-term coverage, though a few common synonyms are missing.	4 / 5
Distinctiveness Conflict Risk	It occupies a clear niche (context-window optimization, lost-in-middle, KV-cache), with distinct triggers; minor overlap risk remains with general summarization/memory skills.	4 / 5
	Total	17 / 20 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
metadata_version	'metadata.version' is missing	Warning
metadata_field	'metadata' should map string keys to string values	Warning

	Total	14 / 16 Passed

Repository: HoangNguyen0403/agent-skills-standard
Path: .github/skills/common/common-context-optimization/SKILL.md
Commit: 405cdd5

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.