administrator/DocForge2
1
0
Fork 0
Code Issues Pull requests Projects Releases 2 Packages Wiki Activity Actions Pages

Guard long-running adapter implementations

Browse source
This commit is contained in:
Andraxion 2026-07-28 19:44:25 -04:00
parent bb13258861
commit 1ef76f0271
11 changed files with 541 additions and 11 deletions
Download patch file Download diff file Expand all files Collapse all files

14
ACTIVE_SLICE.md
View file

@ -1,13 +1,13 @@
# Active slice
```text
Slice: DFG-22 deterministic incremental adapter assembly
Goal: Let language frontends cache repeated raw source evidence while publishing one deterministic graph without maintaining a second project-owned extraction cache.
In scope: Optional generic assembly contract; cached contribution assembly; manifest-bound identity checks; final Logic ownership checks; overlapping-evidence tests; incremental documentation and onboarding guidance.
Out of scope: Language-specific merge rules; bundled C++, Rust, Java, or other frontends; changes to the default unique-ownership path; hidden source reads during assembly; project-owned cache formats; source mutation; deployment or publication.
Done when: An adapter can cache overlapping source contributions, deterministically assemble one valid node, reuse the warm cache, reparse only one changed source, preserve the selected published fact, pass full/incremental equivalence, and the complete DocForge quality gate passes.
Owners: DocForge owns cached contribution delivery, the optional assembly boundary, identity validation, and final graph validation. The adapter owns deterministic evidence selection and merge semantics. Canonical source files remain authoritative.
Proof: The focused adapter and onboarding suite passed 14 tests. Strict Pyright passed with no errors or warnings. Ruff lint and formatting, Python compilation, and the HTML/CSS/JavaScript quality gate passed. The complete warning-strict suite passed 82 tests and 2 subtests. The overlap fixture cached two source contributions, published one deterministic node, reused both warm entries, reparsed only one changed source, preserved the selected node, and passed full/incremental equivalence.
Slice: DFG-23 process-stable adapter implementation boundary
Goal: Prevent a long-running project server from synchronizing with adapter code or configuration that changed after the adapter object was imported.
In scope: Language-neutral implementation roots/files/suffixes; project-local Python package inference; descriptor fingerprinting; bounded change evidence; restart-required MCP remediation; staged/unstaged deletion contract guidance.
Out of scope: In-process Python module reloading; MCP self-restart; project-specific Git enumeration; canonical source mutation; deployment or publication.
Done when: Every MCP operation rejects added, changed, deleted, missing, or unsafe adapter implementation files before synchronization; derived files outside the declared boundary remain ignored; descriptor changes require restart; deletion semantics remain manifest-owned and staging-independent; the complete DocForge quality gate passes.
Owners: DocForge owns implementation-boundary confinement, fingerprinting, bounded diagnostics, MCP preflight, and restart remediation. Adapters own the declared boundary and current source-manifest enumeration. Canonical sources and Git staging remain outside this lifecycle guard.
Proof: The focused adapter suite passed 15 tests, including inferred and explicit implementation boundaries, descriptor changes, additions, edits, deletions, ignored derived files, and exact MCP remediation. Strict Pyright passed with no errors or warnings. Ruff lint and formatting, Python compilation, and the HTML/CSS/JavaScript quality gate passed. The complete warning-strict suite passed 87 tests and 2 subtests.
```
**Next gate:** Prove Worldforge's C++ integration against the generic incremental and assembly