Skip to content

shadow-call-intercept.md

Latest commit

 

History

History
76 lines (56 loc) · 2.91 KB

shadow-call-intercept.md

File metadata and controls

76 lines (56 loc) · 2.91 KB
Raw

Shadow Call Intercept

What are shadow calls?

Codex Desktop App makes background API calls with a hard-coded helper model for internal tasks:

  • Thread title generation — auto-generates 3-8 word titles after your first prompt
  • Commit message generation — generates git commit messages
  • Skill orchestration — internal orchestration turns

These calls happen independently of your selected main model and use reasoningEffort: low.

The helper model is not stable across client versions. Codex used gpt-5.4-mini up to 0.144.x and moved to gpt-5.6-luna in 0.145.0, which silently disabled a single-literal intercept (#311). The intercept therefore matches a set of source-model prefixes — gpt-5.4-mini and gpt-5.6-luna by default — so a client bump does not quietly turn the feature off. Routed ids (provider/model) are never matched: a shadow call is always a bare native slug, and an explicit routed selection must not be hijacked.

The problem

  • Non-OpenAI providers (Bedrock, Azure, etc.) may not support the bare helper slug, causing 404 errors
  • Users have no control over which model handles these helper tasks
  • Shadow calls consume API quota without user awareness

Related GitHub issues: #26288, #28741, #28821, #24208

Configuration

Via Dashboard UI

  1. Open the opencodex dashboard
  2. Find the "Shadow Call Intercept" panel
  3. Toggle the switch to enable
  4. Enter a replacement model (e.g., gpt-5.5)

The panel badge and tooltip render the source models the running proxy actually intercepts, read from GET /api/shadow-call-settings (sourceModels), so the UI never names a stale slug.

Via config.json

{
  "shadowCallIntercept": {
    "enabled": true,
    "model": "gpt-5.5"
  }
}

sourceModels is an optional override. When set to a non-empty array of strings it replaces the defaults rather than extending them:

{
  "shadowCallIntercept": {
    "enabled": true,
    "model": "gpt-5.5",
    "sourceModels": ["gpt-5.6-luna"]
  }
}

Behavior

  • When enabled, ALL requests whose bare model id starts with one of the source-model prefixes (default gpt-5.4-mini, gpt-5.6-luna) are rewritten to the configured model
  • Reasoning effort is forced to low (matching the original behavior)
  • The original model ID is logged as shadowCallRewrittenFrom in request logs
  • When disabled (default), no interception occurs

Warning

Enabling this redirects every request for a source model, not just Codex's background helper turns. gpt-5.6-luna is also a selectable chat model, so if you pick it as your main model while the intercept is on, those turns are redirected too — narrow sourceModels if that matters to you.