OpenCLI AutoFix — Automatic Adapter Self-Repair

When an opencli command fails because a website changed its DOM, API, or response schema, automatically diagnose, fix the adapter, and retry — don't just report the error.

Safety Boundaries

Before starting any repair, check these hard stops:

• AUTH_REQUIRED (exit code 77) — STOP. Do not modify code. Tell the user to log into the site in Chrome.
• BROWSER_CONNECT (exit code 69) — STOP. Do not modify code. Tell the user to run opencli doctor.
• CAPTCHA / rate limiting — STOP. Not an adapter issue.

Scope constraint:

• Only modify the file at adapterSourcePath in the trace summary.md front matter — this is the authoritative adapter location (may be clis/<site>/ in repo or ~/.opencli/clis/<site>/ for npm installs)
• Never modify src/, extension/, tests/, package.json, or tsconfig.json

Retry budget: Max 3 repair rounds per failure. If 3 rounds of diagnose → fix → retry don't resolve it, stop and report what was tried.

Prerequisites

opencli doctor    # Verify extension + daemon connectivity

When to Use This Skill

Use when opencli <site> <command> fails with repairable errors:

• SELECTOR — element not found (DOM changed)
• EMPTY_RESULT — no data returned (API response changed)
• API_ERROR / NETWORK — endpoint moved or broke
• PAGE_CHANGED — page structure no longer matches
• COMMAND_EXEC — runtime error in adapter logic
• TIMEOUT — page loads differently, adapter waits for wrong thing

Before Entering Repair: "Empty" ≠ "Broken"

EMPTY_RESULT — and sometimes a structurally-valid SELECTOR that returns nothing — is often not an adapter bug. Platforms actively degrade results under anti-scrape heuristics, and a "not found" response from the site doesn't mean the content is actually missing. Rule this out before committing to a repair round:

• Retry with an alternative query or entry point. If opencli xiaohongshu search "X" returns 0 but opencli xiaohongshu search "X 攻略" returns 20, the adapter is fine — the platform was shaping results for the first query.
• Spot-check in a normal Chrome tab. If the data is visible in the user's own browser but the adapter comes back empty, the issue is usually authentication state, rate limiting, or a soft block — not a code bug. The fix is opencli doctor / re-login, not editing source.
• Look for soft 404s. Sites like xiaohongshu / weibo / douyin return HTTP 200 with an empty payload instead of a real 404 when an item is hidden or deleted. The snapshot will look structurally correct. A retry 2-3 seconds later often distinguishes "temporarily hidden" from "actually gone".
• "0 results" from a search is an answer. If the adapter successfully reached the search endpoint, got an HTTP 200, and the platform returned results: [], that is a valid answer — report it to the user as "no matches for this query" rather than patching the adapter.

Only proceed to Step 1 if the empty/selector-missing result is reproducible across retries and alternative entry points. Otherwise you're patching a working adapter to chase noise, and the patched version will break the next working path.

Step 1: Collect Trace Context

Run the failing command with failure-retained trace enabled:

opencli <site> <command> [args...] --trace retain-on-failure 2>trace-error.yaml

On failure, stderr contains the normal error envelope plus a small trace block:

ok: false
error:
  code: SELECTOR
  message: "Could not find element: .old-selector"
trace:
  schemaVersion: 1
  opencliVersion: "..."
  traceId: "..."
  dir: "/path/to/.opencli/profiles/default/traces/..."
  summaryPath: "/path/to/.opencli/profiles/default/traces/.../summary.md"
  receiptPath: "/path/to/.opencli/profiles/default/traces/.../receipt.json"

Read summaryPath first. It is the LLM-oriented entry point and includes front matter:

---
schemaVersion: 1
opencliVersion: "..."
traceId: "..."
status: failure
site: "example"
command: "example/search"
adapterSourcePath: "/path/to/clis/example/search.js"
errorCode: "SELECTOR"
errorMessage: "Could not find element: .old-selector"
---

The artifact directory contains:

summary.md      # start here
receipt.json    # machine-readable trace receipt
trace.jsonl     # full redacted timeline
network.jsonl   # redacted network events
console.jsonl   # redacted console events
state/          # final snapshots when available
screenshots/    # final screenshots when available

If you redirected stderr to a file, read that file and copy trace.summaryPath.

Do not ask the user to rerun with legacy diagnostic env vars. Trace is the repair evidence path.

Step 2: Analyze the Failure

Read the trace summary and the adapter source. Classify the root cause:

Error Code	Likely Cause	Repair Strategy
SELECTOR	DOM restructured, class/id renamed	Explore current DOM → find new selector
EMPTY_RESULT	API response schema changed, or data moved	Check network → find new response path
API_ERROR	Endpoint URL changed, new params required	Discover new API via network intercept
AUTH_REQUIRED	Login flow changed, cookies expired	STOP — tell user to log in, do not modify code
TIMEOUT	Page loads differently, spinner/lazy-load	Add/update wait conditions
PAGE_CHANGED	Major redesign	May need full adapter rewrite

Key questions to answer:

1. What is the adapter trying to do? (Read the file at adapterSourcePath)
2. What did the page look like when it failed? (Read summary.md, then state/ if needed)
3. What network requests happened? (Read Failed Network in summary.md, then network.jsonl if needed)
4. What's the gap between what the adapter expects and what the page provides?

Step 3: Explore the Current Website

Use opencli browser to inspect the live website. Never use the broken adapter — it will just fail again.

DOM changed (SELECTOR errors)

# Open the page and inspect current DOM
opencli browser open https://example.com/target-page && opencli browser --session research state

# Look for elements that match the adapter's intent
# Compare the snapshot with what the adapter expects

API changed (API_ERROR, EMPTY_RESULT)

# Open page with network interceptor, then trigger the action manually
opencli browser open https://example.com/target-page && opencli browser --session research state

# Interact to trigger API calls
opencli browser click <N> && opencli browser --session research network

# Narrow to the request you care about by the fields its body should have
opencli browser --session research network --filter author,text,likes

# Inspect specific API response (key is the `key` field from the default JSON output)
opencli browser --session research network --detail <key>

Step 4: Patch the Adapter

Read the adapter source file at adapterSourcePath from the trace summary front matter and make targeted fixes. This path is authoritative — it may be in the repo (clis/) or user-local (~/.opencli/clis/).

Use the Read tool on the exact path from summary.md front matter.

Common Fixes

Selector update:

// Before: page.evaluate('document.querySelector(".old-class")...')
// After:  page.evaluate('document.querySelector(".new-class")...')

API endpoint change:

// Before: const resp = await page.evaluate(`fetch('/api/v1/old-endpoint')...`)
// After:  const resp = await page.evaluate(`fetch('/api/v2/new-endpoint')...`)

Response schema change:

// Before: const items = data.results
// After:  const items = data.data.items  // API now nests under "data"

Wait condition update:

// Before: await page.wait({ selector: '.loading-spinner', hidden: true })
// After:  await page.wait({ selector: '[data-loaded="true"]' })

Rules for Patching

1. Make minimal changes — fix only what's broken, don't refactor
2. Keep the same output structure — columns and return format must stay compatible
3. Prefer API over DOM scraping — if you discover a JSON API during exploration, switch to it
4. Use @jackwener/opencli/* imports only — never add third-party package imports
5. Test after patching — run the command again to verify
6. Never relax verify/<cmd>.json fixtures to silence a failure. A failing patterns / notEmpty / mustNotContain / mustBeTruthy rule means the adapter's output is broken. Tighten the adapter so it produces correct values; do not loosen the fixture to accept the broken values. The one legitimate reason to edit a fixture during repair is when the site itself changed shape (e.g. URL format migration) — in that case update the fixture and note the change in ~/.opencli/sites/<site>/notes.md. Otherwise editing the fixture is covering up a silent correctness regression.

Step 5: Verify the Fix

# Run the command normally
opencli <site> <command> [args...]

If it still fails, go back to Step 1 and collect a fresh trace. You have a budget of 3 repair rounds (trace → fix → retry). If the same error persists after a fix, try a different approach. After 3 rounds, stop and report what was tried.

Step 6: File an Upstream Issue

If the retry passes, the local adapter has drifted from upstream. File a GitHub issue so the fix flows back to jackwener/OpenCLI.

Do NOT file for:

• AUTH_REQUIRED, BROWSER_CONNECT, ARGUMENT, CONFIG — environment/usage issues, not adapter bugs
• CAPTCHA or rate limiting — not fixable upstream
• Failures you couldn't actually fix (3 rounds exhausted)

Only file after a verified local fix — the retry must pass first.

Procedure:

1. Prepare the issue content from the trace summary you already have:
   • Title: [autofix] <site>/<command>: <error_code> (e.g. [autofix] zhihu/hot: SELECTOR)
   • Body (use this template):

## Summary
OpenCLI autofix repaired this adapter locally, and the retry passed.

## Adapter
- Site: `<site>`
- Command: `<command>`
- OpenCLI version: `<version from opencli --version>`

## Original failure
- Error code: `<error_code>`

~~~
<error_message>
~~~

## Local fix summary

~~~
<1-2 sentence description of what you changed and why>
~~~

_Issue filed by OpenCLI autofix after a verified local repair._

2. Ask the user before filing. Show them the draft title and body. Only proceed if they confirm.

3. If the user approves and gh auth status succeeds:

gh issue create --repo jackwener/OpenCLI \
  --title "[autofix] <site>/<command>: <error_code>" \
  --body "<the body above>"

If gh is not installed or not authenticated, tell the user and skip — do not error out.

When to Stop

Hard stops (do not modify code):

• AUTH_REQUIRED / BROWSER_CONNECT — environment issue, not adapter bug
• Site requires CAPTCHA — can't automate this
• Rate limited / IP blocked — not an adapter issue

Soft stops (report after attempting):

• 3 repair rounds exhausted — stop, report what was tried and what failed
• Feature completely removed — the data no longer exists
• Major redesign — needs full adapter rewrite via opencli-adapter-author skill

In all stop cases, clearly communicate the situation to the user rather than making futile patches.

Example Repair Session

1. User runs: opencli zhihu hot
   → Fails: SELECTOR "Could not find element: .HotList-item"

2. AI runs: opencli zhihu hot --trace retain-on-failure 2>trace-error.yaml
   → Gets trace summary with final state and failed action evidence

3. AI reads summary/state: page loaded but uses ".HotItem" instead of ".HotList-item"

4. AI explores: opencli browser open https://www.zhihu.com/hot && opencli browser --session research state
   → Confirms new class name ".HotItem" with child ".HotItem-content"

5. AI patches: Edit adapter at `adapterSourcePath` — replace ".HotList-item" with ".HotItem"

6. AI verifies: opencli zhihu hot
   → Success: returns hot topics

7. AI prepares upstream issue draft, shows it to the user

8. User approves → AI runs: gh issue create --repo jackwener/OpenCLI --title "[autofix] zhihu/hot: SELECTOR" --body "..."