Add support for 'skipped' result in decision log and dashboard snapshot#1
Add support for 'skipped' result in decision log and dashboard snapshot#1
Conversation
![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: 7ed77a506a
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".
| return { | ||
| usdcBalance: BigInt(0), | ||
| totalPortfolioUsd: 0, | ||
| positions: [], | ||
| lastUpdatedAt: new Date().toISOString(), |
There was a problem hiding this comment.
When getVault returns the zero address, this path sets treasury balances to zero instead of failing closed at decision time. In that state, a buy can still be risk-approved with effectiveSizeUsd = 0, and runExecute proceeds until execute throws Resolved amountIn is zero, causing the whole cycle to fail rather than logging a normal blocked/skipped decision. This now impacts new or unfunded users whose vault has not been created yet.
Useful? React with 👍 / 👎.
| if (autopilotSessions.length === 0) { | ||
| await runCycle(config, "cron"); | ||
| } else { | ||
| for (const session of autopilotSessions) { | ||
| await runCycle(config, "cron", session); |
There was a problem hiding this comment.
This branch disables the global cycle as soon as any autopilot session exists, and only runs session-scoped cycles thereafter. Because global state is still what /api/status and the x402 runtime getters expose in this file, those outputs become stale once a single user session is active. Running the global cycle in addition to per-session cycles (or migrating those consumers to session-aware data) would prevent that regression.
Useful? React with 👍 / 👎.
1 participant
No description provided.