alkgen

8 Commits

Author	SHA1	Message	Date
glm-5.3-flash	a52c8a6b37	docs: typed-output direction set — russh-sftp shape, (a) vs (b) fork - branch 1 output contract targets typed Rust structs (no Value materialization): the kernel-fusion move, alkminer precedent - russh-sftp protocol-type shape as target (plain structs, field order = decode order, shared cursor codec core) - fork: (a) serde-derive + generic Deserializer (interop) vs (b) generated direct codecs (zero overhead, the 72ns shape) - follow-up #1 sharpened: typed-output POC renders Packet both ways, benchmarks against the raw-decode baseline	2026-09-09 07:23:56 +00:00
glm-5.3-flash	b86dadcf77	docs: reader POC complete — OQ-GN-05/06/08 resolved with evidence - reader-poc-summary.md: plan-driven handlebars codegen validated end-to-end (parity-exact vs interpreter, ~1.9x faster, 7 tests) - OQ-GN-05 resolved: self-contained generated code (a), uniform cursor addressing (F-1); OQ-GN-06: serde_json-only dependency - decomposition finding: 20x-vs-hand-rolled gap is Value allocation, not dispatch; output-contract ADR is the big Phase 1 decision - F-1..F-5 findings; Phase 1 follow-up list	2026-09-09 07:12:01 +00:00
glm-5.3-flash	79c2794a99	docs: phase-0 — OQ-GN-01 resolved: dev tool, static files, module output - clarification: alkcall is a library others build on; alkgen branches 1+2 are a dev tool generating static files into downstream projects (protocol crates alktty/alktunnels/sftp, cloud integrations vast.ai/ runpod) from provided schemas (BAST or jsonschema/openapi) - OQ-GN-01 resolved: module-level output for v1; fragments out (defeats dev-tool purpose), crate generation out for v1 (composed from modules later if asked). Residue: file partitioning policy - OQ-GN-06 reframed: generated files land in projects that don't depend on alktype today, so dependency footprint is consumer's choice, not wire-shaped; leaning self-contained, POC decides - branch 2 reframed: static counterpart of alkhttp from_* adapters	2026-09-08 08:58:50 +00:00
glm-5.3-flash	0e9bdf65ab	docs: phase-0 — OQ-GN-02 direction set: plan-driven contexts - plans for offsets/sizes (already computed+validated by the engine, free at generation time since alktype is already a dep), BAST for descriptive metadata; drift risk inverts into ordinary dependency-upgrade tracking of alktype's plan API - residue: exact plan->context projection is the Phase 1 context- builder spec task	2026-09-08 08:46:57 +00:00
glm-5.3-flash	ea1ef10bc3	docs: phase-0 — resolve OQ-GN-03 posture, calibrate OQ-GN-08 POC - OQ-GN-03 resolved as posture: interpreter/compiler duplication is the point (dynamic vs static deployment profiles); alktype near schema-method floor, ~20x off hand-rolled; alkgen's value prop is recovering most of that gap. Drift stays the standing risk. - OQ-GN-08 POC 1 calibration: three-way benchmark (plan-walk vs generated vs hand-rolled) doubles as drift check + perf validation; OQ-GN-05 boundary reframed as performance trade	2026-09-08 08:44:52 +00:00
glm-5.3-flash	105d553353	docs: phase-0 findings — prior art survey + OQ-GN-01..08 - settled: stack shape, wasm-clean validated, handlebars 6.4 pin, strict-mode posture, template-is-code, determinism - prior art: typebox-rs codegen, alkminer-v1 ShaderBuilder, alktype plans as implied spec, alkhttp adapters - OQs: output unit, context source (BAST vs plans), interpreter/ compiler duplication cut, delegation boundary, validation strategy, POC scope	2026-09-08 08:27:48 +00:00
glm-5.3-flash	3178621adb	init: repo skeleton — cargo crate, AGENTS.md, gitignore, licenses - Cargo.toml: alkgen 0.1.0, deps alktype 0.3.0 + handlebars 6.4 + serde_json (preserve_order) + thiserror; empty lib - AGENTS.md: family-style conventions (wasm-clean, strict templates, template-is-code, generated-code hygiene) + reference map - verified: cargo check/clippy/fmt, wasm32-unknown-unknown release build	2026-09-08 08:27:03 +00:00
glm-5.3-flash	a73f197786	init	2026-09-08 07:55:53 +00:00