Files
coder/agent/agentcontext/paths.go
T
Kyle Carberry cd3692c0c2 feat: add agent-side workspace context sources and Agent API v2.10 PushContextState (#25983)
Adds the agent half of the workspace context sources RFC. The agent now
resolves instruction files, skills, and MCP configs into a typed
`Snapshot`, watches the relevant paths recursively, exposes the source
list over a workspace-agent HTTP API, and pushes each `Snapshot` to
coderd over a new `PushContextState` RPC on Agent API v2.10.

The coderd-side handler is a stub returning `Unimplemented` for now.
Real persistence to `workspace_agent_context`, chatd hydration on dirty
events, and the `KindMCPServer` MCP provider are tracked by
[CODAGT-569](https://linear.app/codercom/issue/CODAGT-569/enable-agent-api-v210-pushcontextstate-bump-currentminor-wire-coderd).
This matches the pattern used for v2.7 `ReportBoundaryLogs` in
[#21293](https://github.com/coder/coder/pull/21293), which bumped the
version and shipped a stub server so the wire and client could iterate
before the persistence layer landed.

## What ships

### agent/agentcontext (new package)

- `Source`, `Resource` (kinds `instruction_file`, `skill`, `mcp_config`,
`mcp_server` plus reserved `plugin`/`hook`/`subagent`/`command`),
`ResourceStatus`, `Snapshot`, `ComputeAggregateHash`.
- `Manager` owns the in-memory source list, performs the initial resolve
synchronously in `NewManager`, runs a re-resolve/watcher loop in `Run`,
exposes
`AddSource`/`RemoveSource`/`Sources`/`HasSource`/`Snapshot`/`SubscribeChanges`/`Resync`/`SeedSources`/`Close`.
- `Resolver` walks scan roots, classifies recognized files, enforces 64
KiB per-resource, 2 MiB aggregate, and 500-resource caps with
`StatusOversize`/`StatusExcluded`/`StatusUnreadable`/`StatusInvalid`
outcomes, skips `node_modules`/`vendor`/etc., validates symlink targets
stay inside the scan root, stamps `SourcePath` on user-derived
resources, and optionally pulls MCP server tool lists via an
`MCPProvider` interface. MCP config resources ship metadata only (size,
hash) so secrets in env blocks never leave the agent.
- `Watcher` is a recursive `fsnotify` wrapper with a 250 ms debounce,
dynamic arming of newly created directories, and an ENOSPC-tolerant
degraded mode that no-ops further syncs until the manager resyncs
explicitly.
- HTTP API for `GET/POST /sources`, `GET/DELETE /sources/{path}`, `POST
/resync` mounted at `/api/v0/context`.
- `Pusher` interface plus `RunPush` goroutine with exponential backoff
capped at 30 s. `DRPCPusher` adapts the generated `DRPCAgentClient210`
to `Pusher` and translates `drpcerr.Unimplemented` to
`ErrPushUnimplemented` so the push loop exits cleanly when talking to
coderd deployments that have not enabled the real handler.

### agent/proto (v2.10)

- New messages `ContextResource`, `PushContextStateRequest`,
`PushContextStateResponse` and the `PushContextState` RPC on `service
Agent`.
- Generated `DRPCAgentClient210` interface and
`codersdk/agentsdk.Client.ConnectRPC210` / `ConnectRPC210WithRole`.
- `tailnet/proto.CurrentMinor` bumped from `9` to `10`.

### Agent wiring

- `agent.Options.Client` declares both v2.9 and v2.10 connectors;
`run()` dials with `ConnectRPC210WithRole`.
- `apiConnRoutineManager` holds a `DRPCAgentClient210`. Existing v2.8
routines keep their narrower `DRPCAgentClient28` signature thanks to
interface embedding.
- `startAgentAPI210` is the v2.10 counterpart to `startAgentAPI` for
routines that need the new client. The push context state routine uses
it.
- A `contextManager` is constructed in `agent.init()`, seeded from the
existing `CODER_AGENT_EXP_*_DIRS` env vars, started in its own goroutine
under `gracefulCtx`, and closed in `agent.Close`.
- `handleManifest` calls `Manager.SeedSources` for sources rooted at the
manifest directory, then `Resync` after `manifest.Swap`, so the snapshot
reflects the workspace working directory immediately instead of waiting
for the next filesystem event.
- HTTP routes mounted at `/api/v0/context` when the manager is up.

### Coderd stub

`coderd/agentapi/context.go` returns `drpcerr.Unimplemented` for
`PushContextState`. The real handler that persists
`workspace_agent_context` rows, hydrates chats, and emits dirty events
lives in CODAGT-569.

## Tests

24 tests across `agent/agentcontext` cover types, paths, resolver
behavior with file caps, skill containers, MCP secret omission, symlink
target validation, the recursive watcher firing on real fsnotify events,
manager source CRUD / `Resync` / `SeedSources` / `Run` lifetime, the
HTTP API, the DRPC adapter, and the push retry / initial-flag /
unimplemented paths. Passes `go test -race -count=2`.

`TestAgent_ContextStatePushed` boots a full agent against
`agenttest.FakeAgentAPI` (which now records `PushContextState` traffic)
and asserts the seeded `AGENTS.md` appears in a snapshot push with
`schema_version = 1`.

<details>
<summary>Notes for reviewers</summary>

- Source CRUD is workspace-agent-token only; coderd is not in the path
for source mutation.
- Per-resource cap 64 KiB, aggregate 2 MiB, count cap 500; resources
past the cap ship with `StatusExcluded` and an empty payload so the
aggregate hash still detects content edits. MCP-emitted resources
enforce both a per-provider count cap and the aggregate byte cap.
- Symlinks inside the scan root are followed; symlinks pointing outside
(or broken) are rejected with `StatusExcluded` so credentials reachable
via a stray symlink stay off the wire.
- The initial push gates `lifecycle = ready` in the eventual full
design. For this PR the `SeedSources` plus `handleManifest`-driven
`Resync` keeps the snapshot fresh; the live push loop ships now and
DRPCPusher translates the coderd `Unimplemented` stub into a clean exit.
- The `PLUGIN`/`HOOK`/`SUBAGENT`/`COMMAND` kinds are reserved in proto
and Go enums but unused; the Claude Code plugin resolver ships in a
follow-up that does not need a schema migration.
- Two follow-ups remain, both tracked by
[CODAGT-569](https://linear.app/codercom/issue/CODAGT-569/enable-agent-api-v210-pushcontextstate-bump-currentminor-wire-coderd):
(1) the chatd-side handler that persists snapshots and dirties chats;
(2) the `coder exp chat context` CLI command set for
`list`/`show`/`add`/`remove`/`refresh`.

</details>

_This PR was authored by Coder Agents on Kyle Carberry's behalf._
2026-06-08 12:08:40 -07:00

122 lines
3.5 KiB
Go

package agentcontext
import (
"os"
"path/filepath"
"strings"
"golang.org/x/xerrors"
)
// CanonicalizePath produces the canonical form of a user-
// supplied path. The result is absolute, has ~ expanded, has
// path-traversal segments collapsed, and has symlinks resolved
// when the target exists. The path is left lexically clean if
// it does not yet exist (so adding a not-yet-created directory
// remains possible).
//
// CanonicalizePath returns the original input when it is empty.
func CanonicalizePath(raw string) (string, error) {
raw = strings.TrimSpace(raw)
if raw == "" {
return "", xerrors.New("path is empty")
}
// Expand ~ and ~/ prefixes against the current user's home
// directory. Other ~user forms are not supported on
// purpose; the agent runs as a known user.
if raw == "~" || strings.HasPrefix(raw, "~/") {
home, err := os.UserHomeDir()
if err != nil {
return "", xerrors.Errorf("expand home dir: %w", err)
}
if raw == "~" {
raw = home
} else {
raw = filepath.Join(home, raw[2:])
}
}
if !filepath.IsAbs(raw) {
// Fail closed: relative paths could mean different
// things depending on the agent's working directory at
// add-time, so require the caller to absolutize first.
return "", xerrors.Errorf("path %q is not absolute", raw)
}
cleaned := filepath.Clean(raw)
if resolved, err := filepath.EvalSymlinks(cleaned); err == nil {
return resolved, nil
}
return cleaned, nil
}
// ValidateSourcePath enforces the path-validation rules from
// the RFC's Authorization section. It rejects:
//
// - Paths containing ".." segments after expansion.
// - Paths resolving outside the supplied allowedRoots, unless
// allowedRoots is empty (which disables the check).
//
// allowedRoots are canonicalized lazily; missing roots are
// silently skipped so a workspace with no $HOME does not break
// validation for project-relative roots.
func ValidateSourcePath(canonical string, allowedRoots []string) error {
if canonical == "" {
return xerrors.New("path is empty")
}
// filepath.Clean drops "." but leaves ".." when no parent
// is available. Reject defensively.
for _, part := range strings.Split(canonical, string(os.PathSeparator)) {
if part == ".." {
return xerrors.Errorf("path %q contains parent traversal segments", canonical)
}
}
if len(allowedRoots) == 0 {
return nil
}
// Build canonical, deduplicated allowed roots. Missing
// roots (e.g. an unconfigured ~/.claude/) are skipped.
roots := make([]string, 0, len(allowedRoots))
seen := make(map[string]struct{}, len(allowedRoots))
for _, raw := range allowedRoots {
c, err := CanonicalizePath(raw)
if err != nil {
continue
}
if _, ok := seen[c]; ok {
continue
}
seen[c] = struct{}{}
roots = append(roots, c)
}
if len(roots) == 0 {
// All configured roots were invalid; treat as "deny
// everything" so misconfiguration fails closed.
return xerrors.Errorf("path %q is not inside any allowed root", canonical)
}
for _, root := range roots {
if pathHasPrefix(canonical, root) {
return nil
}
}
return xerrors.Errorf("path %q is not inside any allowed root", canonical)
}
// pathHasPrefix reports whether path is equal to or a
// descendant of prefix. Both arguments must already be clean,
// absolute paths.
func pathHasPrefix(path, prefix string) bool {
if path == prefix {
return true
}
withSep := prefix
if !strings.HasSuffix(withSep, string(os.PathSeparator)) {
withSep += string(os.PathSeparator)
}
return strings.HasPrefix(path, withSep)
}