Preserve the original contributor histories and apply the exact independently reviewed repair.
Source-Parent: 81df6475b2
Review-Manifest-SHA256: 272ae1970f0b94fe40dbb974b509b7333682176259f5a1493bbae12d9a7e70cb
The project-scoped reference file shipped an active [profiles.yolo] block
with approval_policy = "never" and exposed MCP servers / approval_policy /
sandbox_mode / notify / profiles.* keys that override the user's own
~/.codex/config.toml when the repo is trusted. AgentShield flags this as
codex-approval-never (CRITICAL) and codex-project-config-escalates (HIGH).
Restrict the project-scoped file to model selection, persistent
instructions, features, and role definitions. Move every other key into a
clearly-marked USER-LEVEL REFERENCE appendix as commented-out examples so
installs still ship a usable template without overriding user policy.
Verified locally with ecc-agentshield scan --path . against .codex/config.toml:
* pre-fix: 2 critical/high findings
* post-fix: 0 critical/high findings
Refs: ARU-10 (ECC-SEC-1.1 review CLEAR), ARU-18 (this PR).
Co-Authored-By: Paperclip <noreply@paperclip.ing>
The global sanity check (check-codex-global-state.sh) hard-fails when
persistent_instructions is missing from ~/.codex/config.toml, but neither
the baseline .codex/config.toml nor the sync script ever define this field.
This causes a clean install to report a failing sanity check even though the
sync otherwise succeeds (#967).
- Add persistent_instructions to the baseline .codex/config.toml so that
users who cp the config get a working default.
- Downgrade the sanity check from fail to warn, since persistent_instructions
is additive and optional — users who rely solely on AGENTS.md should not be
blocked.
Fixes#967 (persistent_instructions part; context7 naming addressed by #970)
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Signed-off-by: Lidang-Jiang <lidangjiang@gmail.com>
On first startup, npx-based MCP servers need to download packages before
they can respond. The default timeout is too short for this, causing
frequent "timed out after 10 seconds" errors reported in #544.
Add startup_timeout_sec = 30 to all five command-based MCP servers
(github, context7, memory, playwright, sequential-thinking). The URL-based
exa server is unaffected and left unchanged.
30 seconds was chosen over the 20s precedent in merge-mcp-config.js to
give extra headroom for slow networks on first run.
Fixes#544
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Signed-off-by: Lidang-Jiang <lidangjiang@gmail.com>
affaan-m