parcadei

explicit-identity

Explicit Identity Across Boundaries

parcadei 3,930 299 Updated 8mo ago
GitHub

Install

npx skillscat add parcadei/continuous-claude-v3/explicit-identity

Install via the SkillsCat registry.

About this skill

This skill enforces the use of explicit identifiers instead of relying on relative terms like "latest" when crossing process or asynchronous boundaries. It prevents race conditions and incorrect data attribution by requiring unique IDs to be passed through entire pipelines. Developers should use this pattern when spawning processes or managing state across distributed systems.

SKILL.md

Explicit Identity Across Boundaries

Never rely on "latest" or "current" when crossing process or async boundaries.

Pattern

Pass explicit identifiers through the entire pipeline. "Most recent" is a race condition.

DO

  • Pass --session-id $ID when spawning processes
  • Store IDs in state files for later correlation
  • Use full UUIDs, not partial matches
  • Keep different ID types separate (don't collapse concepts)

DON'T

  • Query for "most recent session" at execution time
  • Assume the current context will still be current after await/spawn
  • Collapse different ID types:
    • session_id = Claude Code session (human-facing)
    • root_span_id = Braintrust trace (query key)
    • turn_span_id = Braintrust turn within session

Example

// BAD: race condition at session boundaries
spawn('analyzer', ['--learn'])  // defaults to "most recent"

// GOOD: explicit identity
spawn('analyzer', ['--learn', '--session-id', input.session_id])

Source Sessions

  • 1c21e6c8: Defined session_id vs root_span_id distinction
  • 6a9f2d7a: Fixed wrong-session attribution via explicit passing
  • a541f08a: Confirmed pattern prevents race at session boundaries