spec: 钉死 agent 执行面边界(ADR-0018)

ADR-0001/0002/0004 覆盖协作治理到 triggerAgent,但触发后 agent 在执行层
能干什么——读哪些文件、跑什么命令——无 ADR / spec 覆盖。ADR-0017 落地时
采用 bypassPermissions + 全量内置工具,agent 文件/shell 面对宿主无界;
workspace.ts 的 confine() 沙箱意图存在但被 ADR-0017 静默覆盖成死代码。

新增 ADR-0018 + Spec.System.AgentSurface 模块补这一层:
- AgentFileOp(run × path)+ Authorized 谓词(路径必须落在 run 工作区内)
- 与 Lock 正交:Lock 限定并发,Surface 限定波及面,都按 run × project 作用域
- 机制 OPEN(工具包装 / OS 沙箱 / SDK 钩子),契约只钉不变式
- ADR-0017 的 bypassPermissions + workspace.ts 死代码标为偏离契约,对齐为 follow-up
This commit is contained in:
2026-07-08 12:37:37 +08:00
parent 6178664696
commit 5b7fd70124
3 changed files with 184 additions and 2 deletions
@@ -0,0 +1,136 @@
# ADR 0018: The Agent Execution Surface Is Bounded By The Project Workspace
## Status
Accepted. The decision is pinned; alignment of the current Hub
implementation to it is a follow-up surfaced below (constitution rule 4:
implementation diverging from contract must be surfaced, not silently kept).
Builds on **ADR-0001** (project group as collaboration space), **ADR-0002**
(project lock scoped to `run_id`), **ADR-0004** (permission grants govern
`triggerAgent`), **ADR-0007** (the engineering file is a directory tree on
disk), and **ADR-0017** (Claude Code SDK as the agent runtime).
## Context
ADR-0001/0002/0004 together cover *collaboration governance* up to the
moment a teacher `@Claude`s: who may trigger an `AgentRun`, that the run
holds a project-scoped lock, that the group is a long-lived space. They stop
at `triggerAgent`. What the agent may do *after* the trigger — which files it
may read or write, what commands it may run — is not named in any ADR, not
modeled in `spec/`, and not enforced by any contract-level invariant.
The gap is not theoretical. ADR-0017 adopted the Claude Code SDK with
`allowedTools: ["Read", "Write", "Bash", "Glob", "Grep"]` and
`permissionMode: "bypassPermissions"`. Under that configuration the SDK's
built-in tools operate over the whole host filesystem with no confinement, and
`Bash` executes arbitrary shell with no approval gate. The Hub's
`hub/src/agent/workspace.ts` was *intended* to be "the agent's only file
surface" (its module doc says so) and ships a `confine()` path validator that
rejects `..` and absolute escapes. But ADR-0017's cutover left those custom
tools unwired — the SDK's built-in tools replaced them — so `confine()` is
dead code and the boundary it was meant to enforce does not exist at runtime.
Compounding this, the systemd unit's default `SERVICE_USER` is `root`
(`hub/deploy/install_service.sh`), so the unbounded agent runs as root.
This is a real divergence point by the constitution's test: "if unwritten,
will the developer and the agent make different assumptions?" They already
have. ADR-0017 assumed the boundary was an implementation concern; the
`workspace.ts` author assumed it was a contract. The next person to touch the
runner has nothing telling them the boundary must be enforced, and ADR-0017
silently overrode the one that existed.
## Decision
### The agent execution surface is bounded by the run's project workspace
During an `AgentRun`, every file operation the agent performs — read, write,
glob, grep target — must fall **within the run's project workspace directory**
(the on-disk engineering-file tree of ADR-0007). Operations whose target path
escapes that tree are refused. This is a platform core invariant, pinned in
`Spec.System.AgentSurface`.
The boundary is **per-run, anchored to the run's project**, the same scope as
the Lock (ADR-0002). This keeps the two invariants composable:
- `Lock` (ADR-0002) bounds **concurrency** — who may be modifying the project.
- `AgentSurface` (this ADR) bounds **blast radius** — what the agent may touch.
Both scope to `run × project`. Acquiring the lock and entering the surface
happen together at run start; releasing the lock and leaving the surface
happen together at run termination.
### Shell execution is subject to the same boundary, by effect
A shell command the agent runs executes with the workspace as its working
directory (already true for the `cph` subprocess, ADR-0016) and its **file
effects** — files read, written, created, or deleted — are subject to the same
workspace boundary. The agent may invoke **named external tools** whose
binaries live outside the workspace (notably `cph` via `CPH_BIN`); the
boundary governs *file effects*, not *binary location*. Arbitrary
`cat /etc/shadow` or `curl … > /etc/...` is out of bounds regardless of how it
is invoked.
### The Hub process runs as a non-privileged service user
The systemd unit must not run as `root`. The default in
`hub/deploy/install_service.sh` changes from `root` to a dedicated unprivileged
user. This is a deployment invariant, recorded here because the unbounded-agent
risk is compounded by root; it is enforced in the unit, not in `spec/` (it is
not a semantic divergence between developer and agent — it is operational
discipline).
### The mechanism is deliberately left open
How the boundary is *enforced* — path validation inside tool wrappers (the
`workspace.ts` `confine()` shape), an OS-level sandbox (bubblewrap / container
/ chroot), the SDK's own permission hooks, or some combination — is `OPEN`.
The contract pins the **invariant** (operations stay in the workspace), not
the **mechanism**. Any mechanism that upholds the invariant satisfies the
contract; switching mechanisms is not a spec change.
## Consequences
- `Spec.System.AgentSurface` gains an `AgentFileOp` structure and an
`Authorized` predicate mirroring `Memory.McpReadRequest.Authorized`: the op
carries a `run` and a `path`; `Authorized` holds iff the path falls within
the run's workspace, via platform-provided `runWorkspace` and `pathWithin`
(representations `OPEN`).
- `hub/src/agent/workspace.ts`'s `confine()` intent is now contract-backed.
Re-wiring it (or replacing it with an OS sandbox) is **mandatory**, not a
preference. Its current dead-code status under ADR-0017 is a divergence to
be addressed, not silently retained.
- ADR-0017's `bypassPermissions` + unrestricted built-in tools is a
**divergence from this ADR** and is surfaced as such (constitution rule 4).
The reconciliation — restrict the SDK tool set, wrap Bash, or sandbox at the
OS level — is the open mechanism question above. Whichever is chosen must
uphold the `AgentSurface` invariant.
- Inbound file/image delivery (`hub/src/feishu/trigger.ts`) lands files under
`${workspaceDir}/.cph/inbox/`, which is inside the boundary — consistent, no
change to the delivery path.
- The `cph` subprocess (ADR-0016) already runs with `cwd = workspaceDir`
(`hub/src/agent/cph.ts`), so it is inside the boundary by construction.
- Force-release of a stuck lock (ADR-0002 admin override) does not cross the
surface boundary — it is a lock-table operation, not an agent operation.
## Open Questions / Deferred
- **Enforcement mechanism.** Tool-wrapper path validation vs OS-level sandbox
(bubblewrap/container) vs SDK permission hooks vs a combination. The
contract pins the invariant; this picks the mechanism. Must be resolved
before production, since the current mechanism is "none".
- **`bypassPermissions` reconciliation.** Whether the Claude Code SDK can be
configured to confine its built-in Read/Write/Glob/Grep to a subtree and to
gate Bash, or whether the Hub must re-introduce custom tools wrapping
`confine()`, is the concrete implementation question ADR-0017 left open and
this ADR now forces.
- **Shell command vocabulary.** Whether the agent's Bash is restricted to a
whitelist (e.g. `cph`, `ls`, `grep`) or allowed arbitrary commands whose
*effects* are then confined, is a policy choice left to implementation. The
invariant governs effects either way.
- **Per-user rate limiting and token budgets.** Orthogonal to the surface
boundary (this ADR is about *what* the agent may touch, not *how often* it
may run). Deferred to a separate concern.
- **Inbound file size / MIME limits.** The inbox lives inside the boundary,
but unbounded file size is a DoS vector distinct from the boundary
invariant. Deferred.