fix(self-driving): drop replay from the step 4 native sources - #1174
fix(self-driving): drop replay from the step 4 native sources#1174skoob13 wants to merge 1 commit into
Conversation
Step 4's skill no longer enables the retired `session_replay` / `session_analysis_cluster` source, so the prompt and the architecture doc that describe step 4 have to say the same thing. Replay reaches the inbox through step 6c's Replay Vision scanners. Step 3b still turns the Session Replay product on, because scanners need recordings. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01RGaVkph7kqgkSgbTxaonbC
🧙 Wizard CIRun the Wizard CI and test your changes against wizard-workbench example apps by replying with a GitHub comment using one of the following commands: Test all apps:
Test all apps in a directory:
Test an individual app:
Show more apps
Test against a Context Mill branch:
Add Results will be posted here when complete. |
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: ef601d1a41
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| enabled because each surface already has its own pipeline (step 4's native | ||
| source for error tracking, step 6c's scanners for replay) that a scout would |
There was a problem hiding this comment.
Apply this scanner-based rationale to every STEP 6 description: the backbone at line 109 and the table at line 138 still say that session-replay is excluded because it is consumed or covered by native sources, directly contradicting the new statement that replay has no native source. The completed-work note around lines 534–543 repeats the obsolete claim as well. Leaving these active instructions inconsistent can lead the next context-mill change to restore the retired replay source.
Useful? React with 👍 / 👎.
1 participant
Problem
The self-driving run prompt and its architecture doc tell the agent that step 4 enables a replay source. It no longer does.
session_replay/session_analysis_cluster, so the row emits nothing.ARCHITECTURE.mdare the wizard-side statement of the same step, so they drift the moment that lands.Nobody sees this directly. It matters because the agent reads the prompt, and the next person to change step 4 reads the doc.
Changes
ARCHITECTURE.mdlist error tracking and support as the step-3b-derived sources. Replay is out, with a line saying step 6c's scanners are the replay path.ARCHITECTURE.mdno longer justifies itself with step 4's native source for replay. It names step 6c's scanners.No user-visible behavior changes. The prompt edit changes one sentence the agent reads; every step and its order stay as they are.
How did you test this code?
vitest run src/lib/programs/__tests__/self-driving-prompt.test.ts src/lib/programs/self-driving/__tests__/step-keys.test.tspasses, and prettier is clean on the touched file. Those cover step order, step labels, and the 6c section the prompt edit sits beside.No wizard run was executed against a live project.
🤖 Agent context
Autonomy: Human-driven (agent-assisted)
Written by Claude Opus 5 in Claude Code, directed by @skoob13, who asked whether a planned move to Replay Vision had shipped and then to fix it.
Skills invoked:
/writing-pr-descriptions.This is the smallest of three PRs. PostHog/context-mill#370 changes what a run actually does; a PostHog/posthog PR removes the same retired source from the desktop app's sources grid. This one only keeps the wizard's own copy of step 4 true.