~bigbes/sr-ht-spec

sr-ht-spec/mcpsrv/search.go -rw-r--r-- 6.7 KiB
824788ab — Eugene Blikh 2 days ago
mcpsrv: mark /mcp uncacheable, fail closed on origin, split tool errors from faults

Three gaps between this surface and the cov/dolt pattern the siblings settled
on. The write tools are untouched, the per-tool grant scheme is untouched, and
the transport stays stateful.

The endpoint set no Cache-Control and no Vary at all. Every answer here depends
entirely on the credential the request carried and says nothing about it in its
URL, and some of them are the whole approved corpus, so a shared cache was free
to keep one and replay it to the next caller. cache.go is a local copy of
ecore/mcphttp.PrivateCache, byte for byte on the header values so the retrofit
is a delete and an import once that commit is published. It commits the headers
on Write and Flush as well as WriteHeader: the SDK answers a POST with an event
stream that never calls WriteHeader, so a wrapper hooking only that one sets
nothing on the response an agent actually gets, while passing every other header
test. Measured — with only WriteHeader hooked the streamed answer leaves with
the SDK's own `no-cache, no-transform` and no Vary.

Vary names both planes, where dolt.sr.ht names Authorization alone. dolt is
right for dolt: its /mcp is bearer-only, so naming Cookie would promise a cache
a dependency the surface never reads. It is wrong here. authn.Resolver.Resolve
prefers a bearer token when one is present but falls through to
login.UsernameFromRequest when none is, and an owner cookie resolves to
KindOwner — which is exactly what Gate admits. On this service the cookie is
the difference between the whole corpus and a 401. This is the one string a
future mcphttp retrofit has to reconcile between the two services.

An origin with no host in it was a warning and then an unguarded endpoint. The
Host allowlist is the only thing protecting /mcp once the SDK's own rebinding
guard is disabled, so that path turned one unparseable config value into a
silently open endpoint indistinguishable in every functional test from a
correctly guarded one. It is a construction error now. The daemon cannot reach
it either way: service.Config.Validate already refuses to start unless the
origin parses and carries a host.

Errors from below travelled to the agent as tool results carrying their own
text, so a dead git object store and a missing document were the same kind of
answer. A tool result means "the call was understood and the thing you asked
for is not there", so an agent reading one for a store outage concludes the
document does not exist and rewrites a specification around a document that is
perfectly real — and the store's own words reached it. errors.go splits the
two: service.ErrNotFound is a tool result whose sentence is built from the
call's own arguments, and everything else is a jsonrpc protocol error saying
"internal server error" with the detail logged. The old tests asserted the
behaviour being removed — that the agent was shown the words "on fire" — and
are replaced by ones that pin the split in both directions.
c7477607 — Eugene Blikh 10 days ago
authn: accept tokens.sr.ht working tokens beside the agent token

A second agent credential plane, next to the existing one rather than in
place of it. The agent_token table, every agent configured with it, and
the refs rule and provenance requirement around it are untouched; the
local plane is removed in a later phase, not this one.

The resolver tries the instance plane first and falls back to the local
store on exactly two refusals, bearer.ErrInvalid and bearer.ErrNotOurs.
spec's local token has no prefix to discriminate on — it is 32 random
bytes in base64, which is precisely what "did not decode as one of ours"
looks like — so the fallback replaces the shape test bench and cover can
afford. ErrRevoked, ErrForbidden and ErrUnavailable are terminal: a
withdrawn credential must not get a second chance at the old door, and an
unreachable daemon must not silently degrade into the legacy plane.

Grants ride on the principal and are checked where the action is known,
never in the middleware, which runs upstream of the router: spec:propose
in service.Propose, below both write surfaces, and spec:read in each read
surface's gate. /mcp checks per tool rather than at its Gate, because one
endpoint carries both kinds and a surface-wide read grant would refuse a
propose-only token at initialize. Principal.Authorize is a no-op off the
instance plane, which is what keeps the local token working.

The instance plane brings an owner where the local token had none, so a
working token belonging to anybody but [sr.ht] owner-name is refused
rather than admitted as a second identity: Principal.Owner is read by the
provenance committer, the refs rule's principal kind and the coreauth
AuthContext, all written for one human.

StatusFor is the one status table. ErrUnavailable is 503 and never 401 —
reading "I could not ask tokens.sr.ht" as "revoked" would refuse every
live instance token while a daemon that is deliberately off the hot path
restarts.

An instance with no [tokens.sr.ht] section builds no instance plane and
starts anyway, serving its own agent token as before.
e0938e94 — Eugene Blikh 27 days ago
refactor: make the filter-polarity trap inexpressible

service.SpaceFilter meant "empty membership selects nothing" — a newly
created project has no members — while search.Query.Spaces was a bare
[]core.SpaceRef whose empty case meant every space. Passing a project's
members into a query therefore turned an empty project into the whole
corpus: a silent scope inversion, invisible when it happens, and passing
every test written with a non-empty project.

The filter moves to core.SpaceFilter with unexported fields, and
search.Query takes it whole. There is no slice to hand over any more, so
the inversion cannot be written. Its three states are distinct: every
space (EverythingFilter), exactly these — possibly none — (SpacesFilter),
and the zero value, which is neither. Search refuses the zero value
rather than defaulting it, because both plausible defaults are wrong for
one of the two callers that can produce one, and returns no hits for a
filter that selects no space without asking the index.

service.SpaceFilter is now an alias for the core type, so ResolveProject
and its callers keep their names. Tests that built a Query without a
scope now say core.EverythingFilter(), which is what they always meant.
8edca94e — bigbes 27 days ago
feat(mcpsrv): MCP read tools, with Host validation replacing the SDK guard

Adds the Phase 2 read tools (spec_search, spec_read, spec_list) over the
service layer.

Two security fixes came out of building them.

The read plane could serve proposal content. gitx resolves ref names, and
service.resolveRev passed any string through, so rev=proposals/42 made the
READ plane hand back unreviewed text — which would then flow into agent
context as though approved, the single failure this service exists to
prevent. ValidateReadRev now admits only the approved-head sentinel or a
full 40-character object name, at the layer all three surfaces share.
Abbreviations are refused too: one that is unique today can become
ambiguous later, so a pinned revision would silently stop meaning one
thing.

The MCP SDK's DNS-rebinding guard rejects a loopback listener whose Host
is not loopback, which is exactly nginx forwarding to 127.0.0.1 — it would
403 only in production, passing every local test. The SDK offers no
allowlist, so the guard is disabled and replaced by a stricter check: Host
must equal the configured origin, or a loopback name for development. A
rebinding attack carries the attacker's name in Host and fails it. An
unusable origin logs loudly rather than quietly unguarding the endpoint.