skill-timing

Skill Timing

Timing instrumentation for skill execution measurement with microsecond precision and anomaly detection.

Quick Start

# Measure a skill execution with checkpoints
Use the skill-timing skill.

skill_name: rule-reviewer
target_file: rules/200-python-core.md
model: claude-sonnet-45
review_mode: FULL
timing_enabled: true

Output: Timing metadata embedded in output file with duration, checkpoints, token costs, and baseline comparison.

Purpose

Enable comprehensive performance measurement and analysis:

• Wall-clock duration - Microsecond precision from start to end
• Checkpoints - Intermediate timing points for bottleneck analysis
• Token tracking - Input/output token counts with cost estimation
• Anomaly detection - Real-time detection of shortcuts and timeouts
• Baseline comparison - Compare against historical averages
• Cross-analysis - Performance across models, agents, and modes

Use this skill when

✅ Use this skill when:

• Measuring skill execution duration
• Comparing performance across models or agents
• Identifying bottlenecks with checkpoints
• Tracking token consumption and costs
• Detecting potential agent shortcuts (suspiciously fast execution)
• Building performance baselines for CI/CD
• Analyzing historical timing trends

❌ Don't use this skill when:

• The skill execution is trivial (<5 seconds expected)
• You're just testing syntax (not measuring actual performance)
• The skill doesn't produce a file output (timing metadata needs a file to embed in)

Inputs

Required (timing-start)

• skill_name: string — Name of the skill being timed
• target_file: path — Target file path
• model: string — Model slug (e.g., claude-sonnet-45)

Required (timing-end)

• run_id: hex string (16 chars) — From timing-start output
• output_file: path — Output file for metadata embedding
• skill_name: string — Skill name (for recovery if run_id lost)

Optional (timing-end)

• input_tokens: integer (default: none) — Input token count
• output_tokens: integer (default: none) — Output token count
• format: string (default: human) — Output format: human, json, markdown, quiet
• dimension_timings: JSON array (default: none) — Per-dimension timing data (see schema below)
• auto_dimension_timings: flag (default: off) — Derive dimension_timings automatically from dim_<name>_start / dim_<name>_end checkpoint pairs. Explicit dimension_timings wins if both are supplied (with WARNING).
• review_mode: string (default: FULL) — Review mode if applicable