Skip to content

Add provider fallback to the chatbot - #2

Open
Code-delta-support wants to merge 6 commits into
mainfrom
Code-delta-support-patch-1
Open

Code-delta-support wants to merge 6 commits into
mainfrom
Code-delta-support-patch-1

Conversation

@Code-delta-support

Copy link
Copy Markdown
Contributor

Adds a provider-fallback path to the chatbot so a failed primary call retries against the secondary provider. CodeDelta runs on this PR automatically and posts its report below.

Add a fallback mechanism to ask OpenAI and Claude.

(cherry picked from commit 665470b)
@github-actions

github-actions Bot commented Jul 22, 2026 •

Copy link
Copy Markdown

CodeDelta report

Churn: 4124 logical statements (+1874 / −2117 / Δ133); REP_CHURN 0.97
TRUE_CHURN: 3977 authored — the other 147 (4% of the total) came from 4 generated files (lockfiles, minified bundles, codegen output)

⚠️ Build files changed: 31 build/CI/packaging/dependency file(s) in this diff — can alter the shipped artifact without touching application source — 2 install hook(s) (code runs on install)

  • safety-demo/new/setup.py MODIFIED ⚠ install hook
  • safety-demo/new/package.json MODIFIED ⚠ install hook
  • safety-demo/new/native/gateway/Makefile DELETED ⚠ fetches remote
  • agent-scan-demo/Makefile DELETED
  • safety-demo/new/Makefile MODIFIED
  • safety-demo/new/configure.ac MODIFIED
  • safety-demo/new/m4/ax_check_net.m4 MODIFIED
  • safety-demo/old/CMakeLists.txt MODIFIED
  • safety-demo/old/Makefile MODIFIED
  • agent-scan-demo/.github/workflows/autofix.yml DELETED
  • …and 12 more (full list in the agent report)
    Dependency manifests changed (9): agent-scan-demo/web/package-lock.json (deleted), churn-demo/new/package-lock.json (modified), safety-demo/new/go.sum (modified), safety-demo/new/requirements.txt (modified), safety-demo/new/web/package-lock.json (deleted), safety-demo/old/Cargo.lock (modified), safety-demo/old/package.json (modified), safety-demo/old/requirements.txt (modified) +1 more — a separate, quieter signal.
    Agent scan: 15 CRITICAL · 6 HIGH · 34 ELEVATED · 13 endpoint-level · 49 text-level
    Governance: 15 file(s) send data to foreign-hosted models (15 non-allied); 3 file(s) call AI inside a loop (cost risk)
    Agent infrastructure: 18 artifact(s) in the tree — 6 tier-3 (rogue residue / credentials): agent-scan-demo/openclaw/gateway.auth.token, safety-demo/new/agent/openclaw/gateway.auth.token, safety-demo/old/agent/openclaw/gateway.auth.token, agent-scan-demo/openclaw, safety-demo/new/agent/openclaw
    🔑 Committed credentials: 3 key-format match(es) in the tree — agent-scan-demo/openclaw/deploy_key.pem:1, safety-demo/new/agent/openclaw/deploy_key.pem:1, safety-demo/old/agent/openclaw/deploy_key.pem:1 (values redacted; rotate first, then remove)

AI signals are pointers for review, not verdicts. Full report attached to this run.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant