Index365 audit and fix

index365 audit and fix

Run an index365 audit, turn findings into a prioritized plan, apply the fixes that live in THIS repo, and re-run to confirm the score actually moved. The agent fixes; index365 supplies the findings and per-finding remediation. This is not a summary, it fixes.

Prerequisites

• Auth works: index365 doctor (needs runs:write; if 401 or missing scope, run index365-setup: do not work around it).
• You are in the repository that serves the audited site (so affectedUrls map to files).
• A projectId (index365 projects list, or index365-add-project for a new domain).

Workflow (copy this checklist into your reply; check off as you go)

• 1. Confirm project + audit type: index365 projects list --json. AI-Readiness (runs start) by default; Marketing Signal (marketing run) if the user asked about demand/traffic/conversion.
• 2. Run and wait: index365 runs start --project <id> --wait --json (record the exact command tree, projectId, optional target URL, baseline score, runId, and baseline idempotency key). Paid action; pass --idempotency-key so a retry can't double-spend.
• 3. Read report + prioritize: index365 reports context <runId> > .index365/report.json (this CLI prints JSON to stdout; redirect to a file). Delegate ranking to a subagent over the FILE; never paste a full report into context. Prefer each finding's machine-readable agentActions over re-deriving the fix. (This is index365-triage-findings.)
• 4. Apply in-repo fixes: per top finding: index365 findings get --run <runId> <findingId> --json; map affectedUrls → file with Glob/Grep. If the user requested preview or approval before application, show a proposed patch and STOP without editing. Otherwise, the direct audit-and-fix request authorizes one logical Edit; show the actual diff. (This is index365-apply-fix, repeated.) STOP and ask if a fix exceeds the finding, or affectedUrls doesn't resolve to a file; skip out-of-repo findings.
• 5. Re-run and verify: repeat the baseline command tree and target URL with a new stable verification key such as verify-<baselineRunId>-batch-1; reuse that new key only for retries of this post-fix request. Never reuse the baseline key, which would replay the baseline run. Compare the new score/findings against the baseline, matching by stable findingId. A fix that doesn't move the score or clear its finding is not done, re-investigate.
• 6. Report the delta: baseline → new score, the findingIds fixed, applied-but-unmoved (re-investigate these), and out-of-repo findings with each humanUrl so the user can handle them in the dashboard.

Output

Write all run artifacts to .index365/ (git-ignored). Field reference: references/finding-schema.md.

Common mistakes

• Summarizing instead of fixing (that's index365-read-report).
• Pasting the whole report into context instead of reading the file with jq/grep.
• Re-deriving a fix when agentActions already specifies it.
• Calling step 5 done without comparing scores against the baseline.
• Editing files for findings whose affectedUrls are DNS/hosting/infra (not in this repo).