SatoshiAgents🛰 Mission Control⚗️ Code Forge
All engineering findings across the stack
SCANNER: CHECKING
LEDGER: CHECKING
SAFE FIXES: —
PROOF GATE: READY
ACP: READY
CLAUDE CODE: ACTIVE
CODEQL: NOT INSTALLED
MUTMUT: READY / AWAITING TESTS
Findings Today
—
Last scan: —
Chronic Findings
—
Repeating across scans
Safe Auto-Fixes
—
Safe to apply · branch + trust gate enforced
Repair Queue
—
Items awaiting repair
Proof Gate
—
Last run: —
💡
Next Best Action
Loading system state...
Case Ledger — cases
Sev Title File Category Status Repeats Days Chronic Mode Fix?
⏳
Loading case ledger...
Proof Gate — Stage Visibility
import_check
—
Awaiting run
lint
—
Awaiting run
semgrep
—
Awaiting run
tests
—
Awaiting run
hypothesis
No invariants defined yet
mutmut
No critical-module tests yet
finding_regression
—
Awaiting run
trust_regression
—
Awaiting run
Hot Modules — by Finding Density
Run a scan to populate
Hot Modules — Trust/Control Risk
Trust rule findings will appear after scan
Case Detail
Select a finding to inspect
⚡ Safe Auto-Fix Engine
What Safe Auto-Fix Does
Applies deterministic safe transforms only — no logic changes. Every fix is applied on a branch, validated for syntax, then run through the proof gate. Nothing reaches main without passing all gates.
Fix types: add timeout · harden exception · add logging · atomic write · env-var path
Available Safe Fixes
—
Actions
🔧 Tool Readiness
Honest status of every tool in the stack. No fake maturity.
Code Doctor
READY — 1,085+ findings tracked
Semgrep 1.156
READY — reliability + trust rules
Aider 0.86
READY — repo-aware surgical edits
Hypothesis 6.151
ACTIVE — 23 invariants (dedup, heartbeat, health, funding math)
mutmut 3.5
ACTIVE — 31 targeted tests, critical logic mutations caught (fork/segfault on macOS limits full run)
Finding Ledger
READY — chronicity tracking live
Safe Auto-Fix
READY — 641 eligible findings
OpenHands 1.5
INSTALLED — local runtime, server off until needed
SWE-agent 1.1.0
INSTALLED — Mode D active (autonomous issue repair)
CodeQL 2.25.1
ACTIVE — 782 findings (25 errors, 44 warnings) · baseline set 2026-03-30
reviewdog 0.21
ACTIVE — flake8 on tools/infra/ + code_forge/, diff-aware PR annotations
Proof Gate
9 GATES — hypothesis + codeql now active
Mode C: OpenHands 1.5 installed but disabled (zero-LLM background mode, 2026-03-30). Re-enable: mv ~/Library/LaunchAgents/com.openclaw.openhands.plist.DISABLED ~/Library/LaunchAgents/com.openclaw.openhands.plist && launchctl load ...
Mode D: SWE-agent 1.1.0 ✅ INSTALLED — agents/sweagent_venv. Trigger via POST /api/actions/mode-d-repair?finding_id=X&dry_run=false. Uses claude-sonnet-4-5 via OpenRouter.
Proof Gate History
🔒
No repair attempts yet
Run a safe auto-fix or queue a repair to see proof gate results here
Ready