hyperguild#138: internal/mcp/tools_notes.go's frontmatterBlock,
internal/api/frontmatter.go's parseFrontmatter, and
internal/frontmatter/frontmatter.go's split each independently detected a
note's --- fence, converging on the same shape through parallel bug-fixing
across today's #123/#126/#128 sagas. Each carried a subtly different gap:
internal/frontmatter's closing-fence match had no trailing-newline
requirement (a document ending "\n---extra" was treated as validly closed);
internal/mcp needed CRLF/BOM guards bolted on across two separate review
rounds after both silently defeated its own HasPrefix check; internal/api
never gained either fix, so the same encoding-dependent bypass was live
there too, just never triggered by a test.
internal/frontmatter is the leaf package (imports nothing internal), so it
becomes the canonical fence detector: exported Split(doc []byte) (block,
rest []byte, ok bool), BOM-stripped and CRLF-normalized, with the stricter
trailing-newline-required close. internal/api and internal/mcp now call it
instead of re-detecting the fence themselves, keeping only what's genuinely
package-specific (api's ordered fmLine round-trip shape, mcp's map + block-
scalar-fold logic).
Found two real regressions while wiring the callers, caught by each
package's own pre-existing byte-exactness tests going red immediately:
Split's block/rest boundary sits one byte differently than either caller's
prior contract (it keeps the fence-adjacent newline on each side, by design
for Fix's own reconstruction) — both callers now TrimSuffix/TrimPrefix
exactly that one byte to restore their exact prior semantics. Neither
regression would have been visible without those tests already existing.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
7 review rounds closed 8 distinct bypasses across two mechanism families: fence-detection (field-name substring, block-containment, fence-stripping, CRLF, UTF-8 BOM) and value-extraction (folded YAML block scalar hiding an invalid value). A known v1 gap remains in the block-scalar fold's handling of blank-line-interrupted continuations — documented on checkClaimEnvelope, accepted deliberately (narrow, requires an already-privileged caller deliberately constructing a specific malformed-YAML shape) rather than chased past diminishing returns. Real fix (shared, tested, encoding-normalized frontmatter parser) tracked separately as hyperguild#138.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Co-authored-by: Mathias <mthbqv@gmail.com>
brain_edit_root was gated but unobserved: on success it emitted a slog.Warn
and nothing else. That proves a rewrite was allowed, not that one happened —
and brain-sync commits it within 15 minutes authored `brain-sync`, so git
history cannot tell a CLAUDE.md rewrite apart from routine note traffic.
A successful brain_edit_root now emits an ntfy notification and an AuditEntry
through the same sink a capture already uses. notes.Result gains
PrevContentHash so both ends of the change are identifiable.
Path and hashes only in the notification body. Reserved at Internal, not
Confidential, because Confidential makes DegradingSink refuse to record while
loki is down rather than buffer. Observation is best-effort and never fails
the call; a failure to observe logs at ERROR.
Ordinary brain_edit on knowledge//wiki/ is deliberately not audited — git
history already carries the same before/after bytes with a diff.
Closes#97.
Co-authored-by: Mathias <mthbqv@gmail.com>
brain_read returns a note verbatim; brain_edit is str_replace with an
exactly-once match, operating on the tree brain-sync commits from.
An adversarial design review before implementation changed three things:
1. Governance files are not in the model-facing tool. Brain content is
ingested from the outside world and reaches a model's context, and that
model would hold the editor — one poisoned page could rewrite the
CLAUDE.md a .dispatch-allow agent later loads as instructions. And
brain-sync merges with --strategy-option=ours, so the live tree beats a
human's corrective push on the conflicting lines. CLAUDE.md now goes
through brain_edit_root: de-surfaced from tools/list and gated on the
local static principal. The scopes are disjoint — the privileged door
has the smaller reach.
2. Reads needed the same allowlist, and brain_get already lacked it: it
returned verbatim session transcripts under archive/ by path. A test
proved the exposure live before the fix.
3. Canonicalise before allowlisting — knowledge/../.git/hooks/post-commit
would otherwise reach a hook that brain-sync executes in a pod holding
the Gitea push credentials.
Plus overlapping match counting, UTF-8 validation both ways, refusal while
git is mid-operation, flock with the read inside the lock, atomic
same-dir rename, and reindex only when frontmatter is touched.
Refs brain#4.
Co-authored-by: Mathias <mthbqv@gmail.com>
mathiasandClaude Sonnet 5