keshav-three-pass

# Keshav Three-Pass Read one paper in three passes of increasing depth. The outputs accumulate as prose rather than fixed fields; use a different tactic when cross-paper alignment matters more than understanding. ## Orchestration Pattern 1. Call `paper-fetch` with `paper_ref`. Stop on `not_found`. Create `context/papers/<dir>/keshav-three-pass/` on success. 2. Call `first-pass-skim` with `source_path` and `meta_path`. Write `01-first-pass-skim.md`, recording `read_deeper` in frontmatter. 3. If `read_deeper` is false, stop by default. Continue only on explicit caller override and record `gate_overridden: true`. 4. Call `second-pass-grasp` with the paths and `skim_notes`; write `02-second-pass-grasp.md`. 5. Call `third-pass-deep-read` with the paths and `grasp_summary`; write `03-third-pass-deep-read.md`. Do not collapse pass 3 into a recap of pass 2. It must surface implicit assumptions, virtual re-implementation mismatches, and concrete improvements. ## Output Layout ```text context/papers/<timestamp>-<title-slug>/ source.md source.meta.json keshav-three-pass/ 01-first-pass-skim.md 02-second-pass-grasp.md 03-third-pass-deep-read.md ``` Each output carries `sop`, `tactic`, and `written_at` frontmatter. Report the gate outcome, core claim, most consequential implicit assumption, unresolved flags, and all output paths.