Fortify

What the source does

• FilesystemNone observedNo reads or writes in the source
• NetworkNone observedNo outbound calls in the source
• ShellNone observedNo commands run in the source
• DependenciesPinnedNothing floats
• SecretsNone foundNo keys in source

Five audit layers

Gates applied: no behavioural pass.

Audit 0.4.1 · source sha 4e163a7a8095 · 19 Oct 2026Report an issue or request a re-scan

• Publishersharpdeveye GitHub userOwner of the repository this skill is read from
• Source foldersharpdeveye/maestro/source/skills/fortifyThe exact folder we audited
• LicenceMITAs the repository states it
• Commit pinned for this verdict4e163a7a8095A new audit runs when the source changes
• Repository statusNot archivedLast push 173 days ago
• Sandboxed runNot runSkills are read, not executed; layer L4 is skipped

Its description: Use when the workflow lacks error handling, has been failing in production, or needs retry logic, fallback strategies, and circuit breakers.

---
name: fortify
description: "Use when the workflow lacks error handling, has been failing in production, or needs retry logic, fallback strategies, and circuit breakers."
argument-hint: "[target area]"
category: fix
version: 2.0.0
user-invocable: true
---

## MANDATORY PREPARATION

Invoke /agent-workflow — it contains workflow principles, anti-patterns, and the **Context Gathering Protocol**. Follow the protocol before proceeding — if no workflow context exists yet, you MUST run /teach-maestro first.
Consult the guardrails-safety reference in the agent-workflow skill for defense-in-depth patterns and error boundary design.

---

Make the workflow resilient. Every external call will fail eventually — model APIs, tools, databases, third-party services. Fortify ensures the workflow handles failure gracefully.

### Fortification Layers

**Layer 1: Input Validation**

- Validate all inputs before processing
- Return clear error messages for invalid input
- Set size limits on all input fields

**Layer 2: Retry with Backoff**
For transient failures (network errors, rate limits, timeouts):

```yaml
Retry strategy:
  max_retries: 3
  initial_delay: 1s
  backoff_multiplier: 2
  max_delay: 30s
  retryable_errors: [429, 500, 502, 503, 504, TIMEOUT, CONNECTION_ERROR]
  non_retryable_errors: [400, 401, 403, 404]
```

**Layer 3: Fallback Responses**
When retries are exhausted:

- Use a cached previous response (if applicable)
- Use a simpler/cheaper model as fallback
- Return a graceful degradation response
- Escalate to human review

**Layer 4: Circuit Breakers**
When a service is consistently failing:

```yaml
Circuit breaker:
  failure_threshold: 5 consecutive failures
  state: CLOSED → OPEN (after threshold) → HALF_OPEN (after cooldown)
  cooldown: 60 seconds
  half_open_max_requests: 1
```

**Layer 5: Timeout Controls**
Every external call needs a timeout:

- Model API calls: 30-120s depending on task
- Tool executions: 10-60s depending on tool
- Database queries: 5-15s
- Third-party APIs: 10-30s

### Fortification Audit

For each component, verify:

- [ ] Input validation present
- [ ] Retry logic for transient failures
- [ ] Fallback for when retries fail
- [ ] Timeout set
- [ ] Error logged with context
- [ ] User gets a meaningful error (not a stack trace)

### Recommended Next Step

After fortification, run `/evaluate` to verify error handling works under realistic failure scenarios.

**NEVER**:

- Retry non-retryable errors (authentication failures, validation errors)
- Retry without backoff (you'll make the problem worse)
- Swallow errors silently (log and handle, don't ignore)
- Set infinite timeouts (they'll hang forever)
- Skip the fallback (retries exhausted with no fallback = user sees an error)

Excerpt, first 2,749 characters of SKILL.md, verbatim from the audited commit. It is text from someone else’s repository, quoted so you can judge it before your agent reads it.

git clone https://github.com/sharpdeveye/maestro
git -C maestro checkout 4e163a7a8095

Works in

1. 2026-10-19
   ✓SAFE B 89 / 100

   first audit · source 4e163a7a8095

Questions about Fortify

• github.com/sharpdeveye/maestro/tree/main/source/skills/fortify — the audited source at commit 4e163a7a8095 and every finding aboveObserved · 19 Oct 2026

Spotted something wrong? Report a correction · Methodology