Encode the conversation's shared understanding as a Manifest at <scratch>/manifest-{ts}.md (writable scratch path — prefer /tmp/ when writable, else $TMPDIR / %TEMP%). If the transcript lacks shared understanding, invoke manifest-dev:figure-out first; propagate --autonomous when invoked from /auto or /do's amendment path. Pre-flight: babysit (--babysit <pr-url> → references/BABYSIT_MODE.md); amend (args contain a manifest file path → see Amendment below); else fresh.

Encoding discipline. figure-out reaches shared understanding of the problem; /define handles manifest-specific encoding judgment calls — invariant vs process guidance, AC scope and pass threshold, phase ordering (fast vs slow), trade-offs to lock as [T-N]. Surface the load-bearing encoding decisions briefly with a recommended answer before encoding; auto-decide the rest and mark (auto) + matching ASM. The manifest is the acceptance contract — what the user accepts as "I'd ship the outcome of executing this."

Task files. Identify task type; load applicable files from tasks/ per tasks/README.md — their gates and Defaults auto-encode before the interview. For code tasks, also compose tasks/PR_LIFECYCLE.md when origin is on github.com; probe if origin is missing or a github-enterprise host. Per-repo for multi-repo manifests.

Manifest Schema

# Definition: [Title]

## 1. Intent & Context
- **Goal:** [High-level purpose]
- **Mental Model:** [Key concepts to understand]

## 2. Approach (Complex Tasks Only)
*Initial direction, not rigid plan. Expect adjustment when reality diverges.*