feat: add dispatching-coding-agents bundled skill#1166
Open
sarahwooders wants to merge 8 commits intomainfrom
Open
feat: add dispatching-coding-agents bundled skill#1166sarahwooders wants to merge 8 commits intomainfrom
sarahwooders wants to merge 8 commits intomainfrom
Conversation
Teaches agents how to shell out to Claude Code and Codex CLIs as stateless sub-agents, including non-interactive execution, model selection, session resumption, and the history-analyzer subagent. 🐾 Generated with [Letta Code](https://letta.com) Co-Authored-By: Letta <noreply@letta.com>
…g skill Explains how to capture session IDs from Claude Code (--output-format json) and Codex (grep session id from output), where session files live on disk, and how to use the history-analyzer subagent to deep-dive into full session data and persist insights to memory. 🐾 Generated with [Letta Code](https://letta.com) Co-Authored-By: Letta <noreply@letta.com>
…ing skill Show end-to-end flows for both Claude Code and Codex: capture session ID from output, construct the session file path, and pass the concrete path to the history-analyzer subagent. 🐾 Generated with [Letta Code](https://letta.com) Co-Authored-By: Letta <noreply@letta.com>
Killed/timed-out claude -p sessions leave stale approval state that causes "stale approval from interrupted session" errors on subsequent runs. Using --no-session-persistence avoids this entirely since we're using Claude Code as a disposable sub-agent. 🐾 Generated with [Letta Code](https://letta.com) Co-Authored-By: Letta <noreply@letta.com>
…stence We want sessions persisted so history-analyzer can read them later. The stale approval issue is solved by skipping permissions entirely (safe in trusted repos for automated dispatch). 🐾 Generated with [Letta Code](https://letta.com) Co-Authored-By: Letta <noreply@letta.com>
Based on testing: Codex handles large repos more reliably than Claude Code and codex-5.3 is the preferred frontier model for dispatch tasks. 🐾 Generated with [Letta Code](https://letta.com) Co-Authored-By: Letta <noreply@letta.com>
The history-analyzer was using `MEMORY_DIR=[provided in assignment]` which resolved to its own agent's memory directory. Changed to use `$LETTA_PARENT_AGENT_ID` (same pattern as the reflection subagent) so memory files are written to the parent agent's repo. 🐾 Generated with [Letta Code](https://letta.com) Co-Authored-By: Letta <noreply@letta.com>
…r-dispatch Single-session analysis produces overly detailed memory that's better represented by the code. Reflection already captures conversation insights. Reserve history-analyzer for bulk migration during /init. 🐾 Generated with [Letta Code](https://letta.com) Co-Authored-By: Letta <noreply@letta.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
dispatching-coding-agentsthat teaches agents how to shell out to Claude Code (claude) and Codex (codex) CLIs as stateless sub-agentsclaude -p,codex exec), model selection, session resumption, thehistory-analyzersubagent for mining past sessions, and dispatch patterns (parallel research, cross-agent review, etc.)Test plan
node build.jscopies fromsrc/skills/builtin/toskills/)/skill dispatching-coding-agentsand confirm SKILL.md renders correctlyclaude -pandcodex execinvocations from within a Letta Code session🐾 Generated with Letta Code