Name: context-window-management
Rating: 34.4 (1 reviews)
Author: administrakt0r

context-window-management

You're a context engineering specialist who has optimized LLM applications handling millions of conversations. You've seen systems hit token limits, suffer context rot, and lose critical information mid-dialogue.

Quality

30%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

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-window-management/SKILL.md

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

Quality

Content

30%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 sparse, partly broken skeleton: it names capabilities and patterns as headings with one-line hints but provides no executable guidance, no workflow, and a truncated sentence. It reads as an unfinished stub rather than a usable skill.

Suggestions

Add concrete, executable guidance for each pattern — e.g. token-count thresholds that trigger each tier, and the actual summarization/routing steps with example prompts or code.

Fix the truncated 'Your cor' sentence and flesh out the empty Anti-Patterns entries with what to avoid and why.

Replace the placeholder 'When to Use' line with explicit trigger conditions and a short decision procedure (when to summarize vs. retrieve vs. trim).

Dimension	Reasoning	Score
Conciseness	The body is short, but it re-states the frontmatter persona and explains concepts Claude already knows ('serial position effect', 'lost-in-the-middle problem', 'context is a finite resource with diminishing returns'), fitting the 'mostly efficient but includes some unnecessary explanation' anchor.	3 / 5
Actionability	It gives only high-level hints ('Place important content at start and end', 'Summarize by importance, not just recency') with no code, commands, thresholds, or concrete steps, matching the 'minimal concrete guidance; high-level hints but missing specific steps' anchor.	2 / 5
Workflow Clarity	There is no sequenced workflow at all — only lists of capabilities and pattern names — and the body is truncated mid-sentence ('Your cor'), fitting the 'steps missing or incoherent; no sequence' anchor.	1 / 5
Progressive Disclosure	Sections exist (Capabilities, Patterns, Anti-Patterns, Related Skills, When to Use) and no external references are needed, but the structure is weak: a broken sentence, empty anti-pattern entries, and a vague non-trigger 'When to Use', matching the 'some structure but could be better organized' anchor.	3 / 5
	Total	9 / 20 Passed

Description

31%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 is a persona statement rather than a capability/trigger statement: it conveys who the skill is 'for' but not what it concretely does or when to invoke it. Voice is second-person and no 'Use when' guidance is present.

Suggestions

Rewrite in third person stating concrete actions, e.g. 'Summarizes, trims, and routes conversation context to keep LLM calls within token limits; chooses between summarization and retrieval by context size.'

Add an explicit trigger clause: 'Use when a conversation is approaching token limits, losing earlier messages to context rot, or when deciding whether to summarize vs. retrieve context.'

Drop the persona framing ('You're a context engineering specialist who has optimized...') — it adds no actionable signal and consumes the description budget.

Dimension	Reasoning	Score
Specificity	The description lists no concrete actions, only a persona/experience frame ('optimized LLM applications handling millions of conversations', 'seen systems hit token limits'), which is vaguer than even the anchor-2 example; the second-person voice ('You're a context engineering specialist') further lowers it.	1 / 5
Completeness	It offers only a vague 'what' (a persona, not a capability) and no 'when'/'Use when' trigger clause at all, fitting the anchor-2 'has a vague what and no when'.	2 / 5
Trigger Term Quality	'token limits', 'context rot', and 'mid-dialogue' are reasonably natural terms, but it misses common variations and synonyms a user would say ('context window', 'summarize conversation', 'trim context'), matching the anchor-3 example.	3 / 5
Distinctiveness Conflict Risk	The context-window-management niche is somewhat specific, but it overlaps with closely related skills it itself names (prompt-caching, conversation-memory, rag), matching the 'somewhat specific but could still overlap' anchor.	3 / 5
	Total	9 / 20 Passed

Validation

93%

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 — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
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
Path: plugins/AI-Agents-Safe-Coding-Skills/skills/context-window-management/SKILL.md
Commit: 57c135b

Reviewed: 3 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.