Commit d961a1a
ci(abi-drift): wire emit→verify gate per cartridge (standards#92 Phase 2) (#98)
## Summary
**Phase 2** of standards#89 sub-issue 3 / standards#92.
Adds `.github/workflows/abi-drift.yml` — a per-cartridge CI gate that
runs `iseriser abi-emit-manifest` → `iseriser abi-verify` on every PR
or push touching a cartridge's `abi/` or `ffi/` tree.
## Allowlist (starts narrow)
Enforced for: **ssg-mcp, k9iser-mcp**.
80 cartridges in the tree have a paired `Safe*.idr` + `*_ffi.zig`.
Spot-check of 5:
| Cartridge | Result |
|---|---|
| ssg-mcp | ✓ clean |
| k9iser-mcp | ✓ clean |
| 007-mcp | ✗ **REAL DRIFT** — `ToolRisk` enum declared in Idris2 but
absent from Zig |
| airtable-mcp | ✗ false-positive — `Error → err` (Zig reserved-word
workaround; converter doesn't know about `error`) |
| postgresql-mcp | ✗ same false-positive |
So: 1 genuine drift, 2 reserved-word false-positives, ~77 unsurveyed.
The allowlist expands as cartridges are audited individually and the
reserved-word handling is added to `iseriser`.
## Dependency chain (block-on)
| | State | Provides |
|---|---|---|
| iseriser#13 | ✅ MERGED earlier today | `abi-verify` subcommand |
| **iseriser#14** | 🟡 DRAFT (opened earlier this session) |
`abi-emit-manifest` subcommand |
| **this PR (boj-server)** | 🟡 DRAFT | the CI gate that uses both |
This PR's CI will fail with `unknown subcommand 'abi-emit-manifest'`
until iseriser#14 merges. **That is the intended block, not a CI
defect.** Mark this ready for review only after iseriser#14 is in.
## Test plan
- [x] workflow YAML syntax valid (single new file, 123 LOC)
- [x] cartridge discovery shell tested locally against the real tree
(returns the 80 cartridges; allowlist correctly narrows to 2)
- [x] live emit→verify on the 2 allowlisted cartridges confirmed clean
(run via release-mode iseriser, exit 0)
- [ ] blocked: cannot test in CI itself until iseriser#14 lands (the
workflow installs from iseriser main branch)
- [ ] reviewer: confirm the install-from-git approach (vs vendored or
precompiled) is acceptable for the initial Phase 2 shape
- [ ] reviewer: confirm the allowlist approach (vs opt-in marker file in
each cartridge) is acceptable
## Why `Refs` not `Closes`
standards#92 is `major,requirements-target` — joint-close on owner
agreement. Phase 2 is one step of a 4-phase plan; not closing here.
Refs hyperpolymath/standards#92
Refs hyperpolymath/standards#89
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>1 parent f52e9e4 commit d961a1a
1 file changed
Lines changed: 123 additions & 0 deletions
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
| 1 | + | |
| 2 | + | |
| 3 | + | |
| 4 | + | |
| 5 | + | |
| 6 | + | |
| 7 | + | |
| 8 | + | |
| 9 | + | |
| 10 | + | |
| 11 | + | |
| 12 | + | |
| 13 | + | |
| 14 | + | |
| 15 | + | |
| 16 | + | |
| 17 | + | |
| 18 | + | |
| 19 | + | |
| 20 | + | |
| 21 | + | |
| 22 | + | |
| 23 | + | |
| 24 | + | |
| 25 | + | |
| 26 | + | |
| 27 | + | |
| 28 | + | |
| 29 | + | |
| 30 | + | |
| 31 | + | |
| 32 | + | |
| 33 | + | |
| 34 | + | |
| 35 | + | |
| 36 | + | |
| 37 | + | |
| 38 | + | |
| 39 | + | |
| 40 | + | |
| 41 | + | |
| 42 | + | |
| 43 | + | |
| 44 | + | |
| 45 | + | |
| 46 | + | |
| 47 | + | |
| 48 | + | |
| 49 | + | |
| 50 | + | |
| 51 | + | |
| 52 | + | |
| 53 | + | |
| 54 | + | |
| 55 | + | |
| 56 | + | |
| 57 | + | |
| 58 | + | |
| 59 | + | |
| 60 | + | |
| 61 | + | |
| 62 | + | |
| 63 | + | |
| 64 | + | |
| 65 | + | |
| 66 | + | |
| 67 | + | |
| 68 | + | |
| 69 | + | |
| 70 | + | |
| 71 | + | |
| 72 | + | |
| 73 | + | |
| 74 | + | |
| 75 | + | |
| 76 | + | |
| 77 | + | |
| 78 | + | |
| 79 | + | |
| 80 | + | |
| 81 | + | |
| 82 | + | |
| 83 | + | |
| 84 | + | |
| 85 | + | |
| 86 | + | |
| 87 | + | |
| 88 | + | |
| 89 | + | |
| 90 | + | |
| 91 | + | |
| 92 | + | |
| 93 | + | |
| 94 | + | |
| 95 | + | |
| 96 | + | |
| 97 | + | |
| 98 | + | |
| 99 | + | |
| 100 | + | |
| 101 | + | |
| 102 | + | |
| 103 | + | |
| 104 | + | |
| 105 | + | |
| 106 | + | |
| 107 | + | |
| 108 | + | |
| 109 | + | |
| 110 | + | |
| 111 | + | |
| 112 | + | |
| 113 | + | |
| 114 | + | |
| 115 | + | |
| 116 | + | |
| 117 | + | |
| 118 | + | |
| 119 | + | |
| 120 | + | |
| 121 | + | |
| 122 | + | |
| 123 | + | |
0 commit comments