Error Recoverer Skill | Agent Skills

Error Recoverer

Detects, classifies, and recovers from errors during autonomous coding sessions.

Quick Start

Handle Error

from scripts.error_recoverer import ErrorRecoverer

recoverer = ErrorRecoverer(project_dir)
result = await recoverer.handle_error(error, context)

if result.recovered:
    print(f"Recovered via: {result.strategy}")
else:
    print(f"Failed: {result.reason}")

Automatic Recovery

@recoverer.with_recovery
async def risky_operation():
    # Operation that might fail
    pass

Error Recovery Workflow

┌─────────────────────────────────────────────────────────────┐
│                    ERROR RECOVERY FLOW                      │
├─────────────────────────────────────────────────────────────┤
│                                                             │
│  1. DETECT                                                  │
│     ├─ Catch exception                                     │
│     ├─ Parse error message                                 │
│     └─ Extract error context                               │
│                                                             │
│  2. CLASSIFY                                                │
│     ├─ Determine error category                            │
│     ├─ Assess severity level                               │
│     └─ Check if recoverable                                │
│                                                             │
│  3. STRATEGIZE                                              │
│     ├─ Query causal memory for similar errors              │
│     ├─ Select recovery strategy                            │
│     └─ Prepare recovery action                             │
│                                                             │
│  4. RECOVER                                                 │
│     ├─ Execute recovery strategy                           │
│     ├─ Verify recovery success                             │
│     └─ Store error→solution chain                          │
│                                                             │
│  5. ESCALATE (if recovery fails)                           │
│     ├─ Rollback to checkpoint                              │
│     ├─ Create detailed error report                        │
│     └─ Signal for human intervention                       │
│                                                             │
└─────────────────────────────────────────────────────────────┘

Error Categories

| Category | Examples | Recovery Strategy | |----------|----------|-------------------| | Transient | Network timeout, rate limit | Retry with backoff | | Resource | File not found, permission denied | Fix path/permissions | | Syntax | Parse error, invalid JSON | Fix syntax errors | | Logic | Test failure, assertion error | Debug and fix code | | Environment | Missing dependency, version mismatch | Install/update deps | | Unrecoverable | Disk full, OOM | Escalate immediately |

Recovery Strategies

class RecoveryStrategy(Enum):
    RETRY = "retry"              # Simple retry
    RETRY_BACKOFF = "backoff"    # Exponential backoff
    ROLLBACK = "rollback"        # Restore checkpoint
    FIX_AND_RETRY = "fix_retry"  # Apply fix, then retry
    SKIP = "skip"                # Skip and continue
    ESCALATE = "escalate"        # Human intervention

Integration Points

memory-manager: Query/store causal chains
checkpoint-manager: Rollback on failure
coding-agent: Provide fixes for code errors
progress-tracker: Log error metrics

References

references/ERROR-CATEGORIES.md - Error classification
references/RECOVERY-STRATEGIES.md - Strategy details

Scripts

scripts/error_recoverer.py - Core recovery logic
scripts/error_classifier.py - Error classification
scripts/retry_handler.py - Retry with backoff
scripts/recovery_strategies.py - Strategy implementations

Agent Skills: Error Recoverer

Install this agent skill to your local

Skill Files