Devin Foley
67ca8287e8
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work > - The `codex_local` adapter isolates local Codex runs by assigning managed `CODEX_HOME` state per company and per agent > - PR #8272 tightened that isolation, but it left a gap: once `CODEX_HOME` became explicit, the adapter treated it like a user-managed override and skipped auth seeding > - That meant newly isolated agents could launch with no usable `auth.json`, hit OpenAI unauthenticated, and fail with `401 Missing bearer` > - Users who had already persisted one of those broken managed homes could remain stranded even after config changes unless Paperclip repaired the home itself > - This pull request teaches Paperclip to seed managed homes correctly, backfill already-stranded managed homes on startup, and reject credential-less managed homes before they reach the provider > - The benefit is that affected managed `codex_local` agents recover automatically after upgrade and restart, without manual `CODEX_HOME` surgery ## Linked Issues or Issue Description - Fixes #497 - Refs #5028 - Related PR: #8272 - Related PR: #8399 ## What Changed - Distinguished Paperclip-managed `CODEX_HOME` paths from genuine external overrides and always seeded auth into managed homes, even when `CODEX_HOME` is explicit in config. - Wrote API-key-backed `auth.json` files for managed homes when `OPENAI_API_KEY` is configured, otherwise symlinked the shared Codex auth for subscription/OAuth flows. - Added a startup reconciliation pass that backfills already-isolated managed homes created by the broken release so upgrade plus restart repairs stranded agents automatically. - Preserved previously resolved API-key auth when the stored `OPENAI_API_KEY` binding is secret-backed and startup cannot resolve the secret value directly. - Hardened the managed-home preflight to require a credential-bearing `auth.json`, not just file presence, and documented the recovery behavior. - Added regression tests covering managed-home seeding, fail-fast behavior, and server-side startup reconciliation. ## Verification ```bash pnpm exec vitest run packages/adapters/codex-local/src/server/codex-home.test.ts packages/adapters/codex-local/src/server/execute.auth.test.ts server/src/__tests__/codex-auth-reconciliation.test.ts server/src/__tests__/codex-local-execute.test.ts server/src/__tests__/server-startup-feedback-export.test.ts pnpm --filter @paperclipai/adapter-codex-local typecheck pnpm --filter @paperclipai/server typecheck pnpm check:tokens pnpm exec vitest run server/src/__tests__/server-startup-feedback-export.test.ts ``` GitHub Actions `PR` workflow is green on latest head `e5b1e08d4`, including policy, typecheck, test shards, build, e2e, serialized server suites, and canary dry run. Greptile Review is green on latest head with 0 comments added. No UI changes. ## Risks - Startup reconciliation now mutates persisted managed Codex homes at boot. Risk is low because it only touches Paperclip-managed company/agent home paths and no-ops when a home already has usable auth. - Genuine external `CODEX_HOME` overrides remain intentionally self-managed, so those users still own repair steps inside their custom home. - Hosts with neither shared Codex auth nor an explicit per-agent API key now fail earlier with a clearer adapter error instead of surfacing a downstream `401`, which changes timing but not capability. ## Model Used - OpenAI Codex, GPT-5-based coding agent in a local Codex session; exact served model ID/context window were not exposed to the session. Tool use and code execution were enabled. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
82 lines
4.5 KiB
Markdown
82 lines
4.5 KiB
Markdown
---
|
|
title: Codex Local
|
|
summary: OpenAI Codex local adapter setup and configuration
|
|
---
|
|
|
|
The `codex_local` adapter runs OpenAI's Codex CLI locally. It supports session persistence via `previous_response_id` chaining and skills injection through the global Codex skills directory.
|
|
|
|
## Prerequisites
|
|
|
|
- Codex CLI installed (`codex` command available)
|
|
- `OPENAI_API_KEY` set in the environment or agent config
|
|
|
|
## Configuration Fields
|
|
|
|
| Field | Type | Required | Description |
|
|
|-------|------|----------|-------------|
|
|
| `cwd` | string | Yes | Working directory for the agent process (absolute path; created automatically if missing when permissions allow) |
|
|
| `model` | string | No | Model to use |
|
|
| `promptTemplate` | string | No | Prompt used for all runs |
|
|
| `env` | object | No | Environment variables (supports secret refs) |
|
|
| `timeoutSec` | number | No | Process timeout (0 = no timeout) |
|
|
| `graceSec` | number | No | Grace period before force-kill |
|
|
| `fastMode` | boolean | No | Enables Codex Fast mode. Currently supported on `gpt-5.4` only and burns credits faster |
|
|
| `dangerouslyBypassApprovalsAndSandbox` | boolean | No | Skip safety checks (dev only) |
|
|
|
|
## Session Persistence
|
|
|
|
Codex uses `previous_response_id` for session continuity. The adapter serializes and restores this across heartbeats, allowing the agent to maintain conversation context.
|
|
|
|
## Skills Injection
|
|
|
|
The adapter symlinks Paperclip skills into the global Codex skills directory (`~/.codex/skills`). Existing user skills are not overwritten.
|
|
|
|
## Fast Mode
|
|
|
|
When `fastMode` is enabled, Paperclip adds Codex config overrides equivalent to:
|
|
|
|
```sh
|
|
-c 'service_tier="fast"' -c 'features.fast_mode=true'
|
|
```
|
|
|
|
Paperclip currently applies that only when the selected model is `gpt-5.4`. On other models, the toggle is preserved in config but ignored at execution time to avoid unsupported runs.
|
|
|
|
## Managed `CODEX_HOME`
|
|
|
|
When Paperclip is running inside a managed worktree instance (`PAPERCLIP_IN_WORKTREE=true`), the adapter instead uses a worktree-isolated `CODEX_HOME` under the Paperclip instance so Codex skills, sessions, logs, and other runtime state do not leak across checkouts. It seeds that isolated home from the user's main Codex home for shared auth/config continuity.
|
|
|
|
### Per-agent isolation and auth seeding
|
|
|
|
For `codex_local` agents the server isolation guard pins each agent to a per-agent home (`<instance>/companies/<companyId>/agents/<agentId>/codex-home`) and sets `OPENAI_API_KEY=""` so an agent can never spend against the host API key or share another agent's Codex state.
|
|
|
|
A managed home is created empty, so the adapter must provision auth into it before launching Codex — otherwise the agent runs with zero credentials and the provider returns `401 Missing bearer`. The seeding contract:
|
|
|
|
- **Managed homes** (the default home and any configured `CODEX_HOME` under the company tree) are always seeded: the ChatGPT-subscription `auth.json` is symlinked from the host Codex home, or, when a per-agent `OPENAI_API_KEY` is configured, an API-key `auth.json` is written instead.
|
|
- **Genuine external overrides** (a `CODEX_HOME` outside the Paperclip-managed company tree) are treated as self-managed and are never seeded or overwritten.
|
|
- **Fail-fast guard:** if a managed home ends up with no usable `auth.json` and no configured API key, the run fails with an explicit `adapter_failed` ("no Codex credentials provisioned for managed home …") rather than emitting an unauthenticated request.
|
|
|
|
## Manual Local CLI
|
|
|
|
For manual local CLI usage outside heartbeat runs (for example running as `codexcoder` directly), use:
|
|
|
|
```sh
|
|
pnpm paperclipai agent local-cli codexcoder --company-id <company-id>
|
|
```
|
|
|
|
This installs any missing skills, creates an agent API key, and prints shell exports to run as that agent.
|
|
|
|
## Instructions Resolution
|
|
|
|
If `instructionsFilePath` is configured, Paperclip reads that file and prepends it to the stdin prompt sent to `codex exec` on every run.
|
|
|
|
This is separate from any workspace-level instruction discovery that Codex itself performs in the run `cwd`. Paperclip does not disable Codex-native repo instruction files, so a repo-local `AGENTS.md` may still be loaded by Codex in addition to the Paperclip-managed agent instructions.
|
|
|
|
## Environment Test
|
|
|
|
The environment test checks:
|
|
|
|
- Codex CLI is installed and accessible
|
|
- Working directory is absolute and available (auto-created if missing and permitted)
|
|
- Authentication signal (`OPENAI_API_KEY` presence)
|
|
- A live hello probe (`codex exec --json -` with prompt `Respond with hello.`) to verify the CLI can actually run
|