Closes#931.
The transport was a literal passed once at the bottom of the module,
bind_native_mcp_transport(transport="stdio"), and every guard that asks
"is this a trusted native session" resolves that question through the
value bound there. The string was a constant in the authorization chain
rather than configuration, so a remote transport had no way to be
expressed and no way to be told apart from an untrusted offline import.
mcp_transport_config becomes the single source of truth: it owns the
permitted set, the default, and the resolution of deployment
configuration (GITEA_MCP_TRANSPORT) into a validated identifier. Unset
configuration still yields stdio, so existing deployments are unchanged.
streamable-http is accepted so the bind is pluggable; standing up its
listener remains #938. sse is deliberately not registered.
bind_native_mcp_transport now resolves from configuration when no
transport argument is given, validates against that one permitted set
before writing the runtime record, and pins the result. bound_transport
is the shared accessor every transport-aware guard reads; it reports the
pinned value and never the environment, so a post-bind GITEA_MCP_TRANSPORT
change cannot move what a guard observes -- the rule already applied to
the session-state root under #695 AC2. Rebinding the same identifier is
idempotent; rebinding a different one is refused, so two guards can never
disagree within one process. assert_transport_bound fails closed before
mcp.run, so an absent or invalid bind stops the server instead of serving
tools over a transport no guard can name.
The bound identifier is now recorded in durable provenance:
mutation_provenance_fields gains bound_transport, which reaches the
decision lock through mcp_session_state.save_state and the audit records
that already spread those fields. The pre-existing transport field keeps
its trust-class values, so no durable record changes shape.
assess_transport_for_auth_mint reports the bound transport through the
same accessor; its verdict is unchanged.
#956's anchor fixture and threat model are re-anchored for the lines this
change shifts, and the two boundary claims that #931 makes false -- the
"literal stdio bind" in B1 and the stdio contract at mcp_server.py:4 --
are restated. Without this the #956 guard fails, which is what it is for.
Non-goals, untouched: the remote listener (#938), per-request principal
resolution (#932), client-managed provenance policy (#934), TLS and
remote client authentication, role capability sets, repository-binding
semantics.
Tests: tests/test_issue_931_transport_bind_seam.py, 42 tests, covering
default bind, explicit stdio, the accepted non-stdio identifier, an
unregistered identifier, no bind at all, one-value agreement across
guards, post-bind environment tampering, the durable decision-lock
record, the rebind contract, and the #695 protections that must not
regress.
Full suite from the branches/ worktree: 28 failed, 5843 passed, 6
skipped, 1047 subtests. The pinned base 9b80e75c reports 28 failed, 5801
passed, 6 skipped, 1047 subtests. The failing test IDs are identical sets
-- no failure added, none fixed; the +42 passed are this PR's new module.
Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Address formal review 438 REQUEST_CHANGES on PR #710 (F6/F7/F8).
F6 — HMAC key-version validation fails closed:
- verify_authorization_artifact validates key_version BEFORE any MAC work, so an
attacker-chosen version can never select the signing key.
- Require exactly one nonempty, well-formed key-version field; missing, empty,
unknown, malformed, duplicated (including identical-valued and nested
aliases), and mismatched versions all fail. No versionless legacy fallback.
- Artifact version must equal the configured active version; production now
requires GITEA_IRRECOVERABLE_AUTH_HMAC_KEY_VERSION explicitly (an implicit
default made rotation ambiguous). Version stays inside the signed material.
F7 — strictly canonical incident evidence:
- Replace substring/first-match parsing with an exact schema: marker on line 1,
every field once, fixed order, no duplicate/unknown/empty/conflicting fields
in or outside the signed block. The parsed body is re-rendered and compared
for exact equality before acceptance.
- content_digest now binds the full recovery scope: repository identity, PR,
decision-lock identity, destroyed subject, recovery action, recorded and
expected head, incident issue, evidence author, minting actor, key version,
nonce and issued_at.
- Actor identity is the immutable user id with login consistency; conflicting
ids/logins and display-name-only identities fail closed. Edited comments are
rejected. The independent-author rule is preserved and enforced by stable id.
- build_canonical_incident_body is the single source of the accepted format and
refuses to emit ambiguous evidence.
F8 — archival is a prerequisite for clearing terminal evidence:
- _clear_decision_lock_for_profile no longer swallows archive failures. It
requires a successful write plus a durable read-back matching the PR/head,
and otherwise returns a structured, retry-safe failure that retains the lock
and records actionable recovery evidence.
- Fixes a latent bug the read-back exposed: the archive payload inherited the
source lock's session_profile_lock, so save_state keyed the archive under the
reviewer profile instead of the archive identity and it never read back.
Adversarial regressions added for every listed case: key-version missing/empty/
unknown/malformed/duplicate/rotation/wrong-key-after-restart, reordered fields,
duplicate identical and conflicting fields, conflicting actor ids/names,
digest-preserving substitution, decision-lock and recovery-action substitution,
cross-PR/repo/org/remote/head replay, archive exception/timeout/false/empty/
partial-readback with proof the lock survives, exactly-one permitted clear, and
retry after archive failure. No existing assertion was weakened.
Validation: focused 150 passed in three module orders; full tests/ 2809 passed,
6 skipped, 1 warning (pre-existing StarletteDeprecationWarning in
tests/test_webui_audit.py:8), 161 subtests passed.
Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
jcwalker3andClaude Opus 4.8
sysadminandClaude Opus 4.8
