Files
archie/docs/prds/config-example-drift.md
T
sam 641c80f7d9 feat(build): gate generated contract data in task check
Closes #855 (archie-core-5gzx).

The check mode added in the preceding commit had no caller, so it prevented nothing: a
stale contracts.json still passed every gate. This wires it in and makes
drift fail the build.

- `task docs:generate` writes the committed artifact.
- `task docs:check` verifies it, non-destructively.
- `task check` runs `docs:check`, right after `proto:check`, so both
  generated-artifact surfaces are judged the same way in the same place.

The check cannot be satisfied by rewriting: it generates into a temporary
directory, so a failing gate leaves the artifact exactly as it found it and
the author applies the regenerated bytes. That keeps the regenerated diff
reviewable in the same commit as the code change that caused it, which is
what Step 9 of the architecture document requires and what a CI auto-update
would quietly bypass.

Documents that called this gap open are updated rather than left to
contradict the code:

- `generated-documentation.md` Step 8 (no `docs:*` task exists -> both exist),
  Step 9 (drift is gated locally, not in CI), and the hook/CI policy (rule 3
  is now enforced by the gate; rules 1, 2, and 4 still rely on the author,
  since nothing observes a commit).
- `contract-conformance-audit.md` gap 1, now closed.
- `config-example-drift.md` item 3, which asked for exactly this wiring.
- `CLAUDE.md`, whose `task check` description listed neither `docs:check`
  nor `test:tools`.
- The build, docs, and diagnostics skills, plus a re-baselined diagnostics
  snapshot: the `task_docs` delivery surface went 0 -> 6 and the `tools`
  module 4 -> 7 now that the tasks exist.

Verified with `task check` (green), including the new `docs:check` step.
2026-09-18 12:46:40 +10:00

124 lines
6.2 KiB
Markdown

# Example-config and generated-config drift — decision
**Status:** Decided, not yet implemented
**Date:** 2026-09-15
**Closes:** the checked-in-config half of audit findings `config-1` and
`repo-root-deployment-2`; strengthens `doc-drift` findings generally.
**Authority:** extends `docs/architecture/generated-documentation.md` and
`docs/prds/config-schema.md`; does not replace either.
## Question
Should `config.example.toml` and `deployments/*.toml` be *generated*, so they
cannot drift from the config code?
## Decision
**No — do not generate them. Gate them instead.** Neither file is a projection
of the schema, and generating either would destroy the thing that makes it
useful. Three verified reasons, in order of weight:
1. **`config.example.toml`'s value is its comments.** It documents **137 keys**
(measured: active *and* commented-out), and those hand-written comments are
the file's purpose — `docs/architecture/configuration.md` states outright
that "`config.example.toml` carries documentation users are meant to
hand-edit". TOML round-tripping does not preserve comments, which is exactly
why `tomlwrite` does line-oriented edits rather than decode/encode. A
generator would emit a correct file with the documentation deleted.
2. **`docs/prds/config-schema.md` forbids building on `config.Config`.** That PRD
rejects "a reflection-based schema over `config.Config`" because the struct is
"scheduled for dissolution". A generator deriving the example from its struct
tags would build on a foundation this repo has already decided to remove.
3. **`docs/architecture/generated-documentation.md:225` prohibits it.** "Adding
another hardcoded type list to `docsgen` is prohibited" — so any generator
must consume a domain-owned registry, which does not exist for configuration
yet. Creating one is the config-domain migration's job, not this change's.
The deployment profiles are also not a projection: they set **3-46 keys** against
the example's 137, and they are curated overlays ("a complete config, not an
overlay", per their own headers), not near-duplicates. Generating them would add
machinery to remove duplication that largely is not there.
## The gap that is real
**Unknown config keys are silently ignored.** Verified at
`internal/infrastructure/configuration/decode.go:27`:
```go
if _, err := toml.DecodeFile(path, target); err != nil {
```
There is no `MetaData.Undecoded()` check, so a typo — `max_concurrancy = 4` —
parses clean, validates clean, and does nothing. That is the same failure shape
this repo's audit found repeatedly: a setting that looks applied and is not. It
is also detectable **without any hand-maintained list**, because `BurntSushi/toml`
reports exactly which keys the target did not consume.
Note the second, separate gap already admitted in writing:
`generated-documentation.md:331` — "`task check` does not run `docs:check`, so
generated drift is currently ungated."
## What to build
**1. Unknown-key detection at the loader (behaviour).** Decode with metadata and
collect `Undecoded()`. Report it as a warning at startup listing the offending
keys, and expose it so tests can assert emptiness.
**Deliberately a warning, not a fatal error.** `configuration.md`'s settled
startup policy is "a missing credential disables a capability; only an invalid
config stops the daemon". A stray key is far more likely to be an operator typo
in a file they are otherwise happy with than a reason to refuse to boot, and
`[web]` is already documented as landing in a *different* binary
(`internal/app/archieui`), so a key unknown to one process is legitimately known
to another. Failing closed here would break working deployments on upgrade.
**2. A checked-in-config contract test (gate).** Load `config.example.toml` and
every `deployments/*.toml` through the real loader and assert `Undecoded()` is
empty, with an explicit allowlist for keys another binary owns. This is the
mechanism that would have caught a renamed key left behind in an example, and it
needs no inventory.
**3. Gate generated data in `task check`.** Done. `task docs:check` runs in the
gate, closing the admitted "generated drift is currently ungated" hole. The
check mode did not exist when this was written; it was added for issue #855,
which also made the check non-destructive so the gate can never rewrite the
artifact it is judging.
**4. Explicitly NOT now: a generated key inventory.** `configuration.md`'s
completion criteria want "generated documentation lists every external key,
default, constraint, mutability class, owner". That is a real target and it is
the config domain's, once a domain-owned registry exists — the prohibition above
means it cannot be bootstrapped from a list inside `docsgen`. Recorded as
deferred, not rejected.
## Non-goals
- No generation of `config.example.toml` or `deployments/*.toml`.
- No reflection-based schema over `config.Config`.
- No new type list inside `docsgen`.
- No change to `archied setup`'s template: `configtemplate.Example` stays the
verbatim embedded example, hand-maintained, and `setup` continues to *edit* it
via `tomlwrite` rather than regenerate it.
## A withdrawn premise, recorded so it is not reused
An earlier measurement in this session claimed 17 keys were "set by a profile
but documented by no example", and used it to argue for generation. It was an
artifact: the extractor stripped comment markers and tracked a single table
variable, so commented blocks were mis-attributed. Re-checking refuted it —
`providers.<name>.api_key_env` is a real, documented backwards-compatible field
(`internal/config/config.go:180`, `config.example.toml:101`) and `[[identities]]`
is documented at `config.example.toml:53`. The decision above does not rest on
that measurement.
## Acceptance criteria
1. A config file containing an unknown key surfaces those keys to the operator at
startup, and the daemon still starts.
2. `config.example.toml` and every `deployments/*.toml` are loaded by a test that
fails if any key is undecoded, except keys explicitly attributed to another
binary.
3. `task check` fails when committed generated data is stale.
4. Introducing a renamed key into `config.example.toml` fails the gate (prove it
by doing it, not by asserting it).