cc-skill-strategic-compact

Development skill from everything-claude-code

1.31x

Quality

Does it follow best practices?

Impact

95%

1.31x

Average score across 3 eval scenarios

Securityby

Passed

No known issues

Optimize this skill with Tessl

npx tessl skill review --optimize ./skills/cc-skill-strategic-compact/SKILL.md

Evaluation results

100%

22%

Context Monitor Script for AI Development Sessions

Tool call counter script implementation

Criteria

Without context

With context

Counter file in /tmp

100%

PID-scoped counter file

100%

COMPACT_THRESHOLD env var

100%

Default threshold = 50

100%

Suggest at threshold

100%

Periodic reminder interval = 25

100%

Output to stderr

100%

Stdout stays clean

100%

Counter increments per call

100%

Counter initializes at 1

100%

Without context: $0.3517 · 1m 47s · 13 turns · 62 in / 6,284 out tokens

With context: $0.3518 · 1m 17s · 19 turns · 325 in / 3,999 out tokens

85%

23%

Automate Context Compaction Reminders in Claude Code

Claude Code hook configuration

Criteria

Without context

With context

PreToolUse hook type

100%

Edit|Write matcher

100%

Hook type = command

100%

Script path correct

Valid JSON structure

100%

PreToolUse rationale documented

100%

Correct trigger tools documented

46%

100%

Without context: $0.2366 · 1m 6s · 10 turns · 11 in / 3,233 out tokens

With context: $0.4519 · 1m 46s · 23 turns · 90 in / 5,560 out tokens

100%

24%

Developer Guide: Context Management in Long Claude Code Sessions

Strategic compaction timing criteria

Criteria

Without context

With context

Manual vs auto rationale

66%

100%

Logical phase preservation

90%

100%

Exploration-to-execution transition

93%

100%

Milestone transition point

93%

100%

Extended session duration criterion

20%

100%

High tool call volume criterion

20%

100%

Plan finalized criterion

92%

100%

No advocacy for auto-compact

100%

Without context: $0.4610 · 1m 55s · 22 turns · 22 in / 5,457 out tokens

With context: $0.4355 · 2m 1s · 24 turns · 20 in / 5,547 out tokens

Repository: sickn33/antigravity-awesome-skills
Commit: 9c177eb

Evaluated: 19 days ago
Agent: Claude Code
Model: Claude Sonnet 4.6

Table of Contents

Context Monitor Script for AI Development Sessions Automate Context Compaction Reminders in Claude Code Developer Guide: Context Management in Long Claude Code Sessions

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.