Secrets

You are in AUTONOMOUS MODE. Do NOT ask questions. Do NOT pause for confirmation. Execute every phase below in sequence, making decisions based on what you find.

============================================================ PHASE 0 — INPUT

$ARGUMENTS may contain:

• A secrets provider: vault, aws-sm (Secrets Manager), aws-ssm (Parameter Store), gcp-sm, doppler, infisical
• --audit — only audit current secret handling, do not generate config
• --rotate — set up secret rotation for database and API credentials
• --ci — configure CI/CD pipeline to pull secrets from provider
• --env-template — generate .env.example from detected env vars
• If no arguments, perform audit and recommend a provider based on existing infrastructure

============================================================ PHASE 1 — SECRET AUDIT

Perform a comprehensive scan for secret exposure:

1. Hardcoded secrets in source code (exclude node_modules, vendor, .git):

• API keys: (api[_-]?key|apikey)\s*[:=]\s*['"][A-Za-z0-9]{16,}['"]
• AWS keys: AKIA[0-9A-Z]{16}
• Private keys: -----BEGIN (RSA|EC|OPENSSH) PRIVATE KEY-----
• Tokens: (token|secret|password)\s*[:=]\s*['"][^'"]{8,}['"]
• Connection strings: (postgres|mysql|mongodb|redis)://[^@]+@
• JWT secrets: (jwt[_-]?secret|signing[_-]?key)\s*[:=]\s*['"]
• Generic high-entropy strings in assignment context (16+ chars, mixed case + digits)

2. Environment files:

• Check for .env, .env.local, .env.production, .env.development
• Check if .env is in .gitignore
• Check if any .env files are tracked by git: git ls-files .env*
• Read .env files and categorize values as: secret vs configuration

3. Git history (last 50 commits):

• Scan diffs for accidentally committed secrets
• Check if any .env files were ever committed then removed
• Flag any file in history matching secret patterns

4. Configuration files:

• docker-compose.yml — passwords in environment blocks
• application.yml / application.properties — embedded credentials
• Terraform files — hardcoded provider credentials or database passwords
• Kubernetes secrets — base64-encoded values in committed YAML

5. CI/CD configuration:

• .github/workflows/*.yml — check for inline secrets (should use ${{ secrets.NAME }})
• .gitlab-ci.yml — check for variables with exposed values

Severity classification:

• CRITICAL: Secrets in source code or git history
• HIGH: .env files tracked by git, secrets in docker-compose
• MEDIUM: .env not in .gitignore, missing .env.example
• LOW: No secrets manager configured, no rotation policy

============================================================ PHASE 2 — ENVIRONMENT TEMPLATE

Generate .env.example from all detected environment variable references.

Scan for env var patterns:

• Node.js: process.env.VARNAME
• Python: os.environ["VARNAME"], os.getenv("VARNAME")
• Go: os.Getenv("VARNAME")
• Generic: ${VARNAME} in config files

Create .env.example with:

• Grouped sections (Application, Database, Cache, Authentication, External APIs, Cloud)
• # SECRET: prefix for values that must come from secrets manager
• # CONFIG: prefix for non-sensitive configuration
• Placeholder values that clearly indicate they need replacement (e.g., REPLACE_ME_WITH_SECURE_RANDOM_VALUE)

============================================================ PHASE 3 — SECRETS PROVIDER SETUP

Based on detected infrastructure or $ARGUMENTS, set up the appropriate provider:

AWS Secrets Manager:

• Generate Terraform for aws_secretsmanager_secret resources
• Generate application helper to read secrets at startup (using @aws-sdk/client-secrets-manager or equivalent)
• Configure IAM policy for ECS task role / Lambda execution role
• Set up secret versioning and stage labels

AWS Systems Manager Parameter Store:

• Generate Terraform for aws_ssm_parameter resources (SecureString type)
• Hierarchical naming: /{project}/{env}/{secret_name}
• Cheaper alternative for smaller number of secrets

GCP Secret Manager:

• Generate Terraform for google_secret_manager_secret resources
• Configure IAM bindings for service account access

HashiCorp Vault:

• Generate Vault policy file and AppRole auth method config
• Docker Compose service for local Vault dev server
• Application helper for Vault API integration

Doppler:

• Generate doppler.yaml project config
• Set up environment configs (dev, staging, prod)

For all providers, generate a unified secrets loading wrapper:

• In production: load from secrets manager
• In development: load from .env file via dotenv

============================================================ PHASE 4 — SECRET ROTATION (if --rotate)

Database credentials:

• AWS: Lambda rotation function with aws_secretsmanager_secret_rotation
• Dual-user rotation strategy (alternating credentials for zero-downtime)
• Rotation schedule: every 30 days

API keys:

• Script that: creates new key, updates secrets manager, waits for propagation, revokes old key
• Document manual rotation steps for third-party APIs

TLS certificates:

• Reference cert-manager or ACM for auto-renewal
• Alert 30 days before expiration

============================================================ PHASE 5 — CI/CD INTEGRATION (if --ci)

GitHub Actions:

• OIDC federation for GitHub -> AWS (no long-lived credentials needed)
• aws-actions/configure-aws-credentials@v4 with role-to-assume
• aws-actions/aws-secretsmanager-get-secrets@v2 for loading secrets
• Environment protection rules for production secrets

Doppler:

• dopplerhq/secrets-fetch-action@v1 with DOPPLER_TOKEN secret

General:

• Enable GitHub secret scanning and push protection
• Document all required repository secrets and where to obtain them

============================================================ PHASE 6 — REMEDIATION

For each CRITICAL and HIGH finding from the audit:

1. Hardcoded secrets: Replace with environment variable references
2. Committed .env files: Add to .gitignore, remove from git tracking (warn about history)
3. Docker compose passwords: Replace with variable references ${DB_PASSWORD}
4. Missing .gitignore entries: Add .env*, *.pem, *.key, *.jks patterns

If secrets were found in git history:

• Recommend git-filter-repo to remove from history (document command but do NOT execute)
• Recommend rotating ALL exposed credentials immediately

============================================================ SELF-HEALING VALIDATION (max 2 iterations)

After completing deployment/infrastructure changes, validate:

1. Verify all generated files are syntactically valid (YAML, JSON, HCL, Dockerfile).
2. Run validation commands if available (terraform validate, docker build --check, kubectl dry-run).
3. Verify no secrets, credentials, or sensitive values are hardcoded.
4. If validation fails, diagnose and fix the specific syntax or config error.
5. Repeat up to 2 iterations.

IF STILL FAILING after 2 iterations:

• Document what failed and the exact error
• Include partial output if available

============================================================ OUTPUT

## Secrets Audit & Configuration

### Audit Results
| Severity | Finding | Location | Status |
|----------|---------|----------|--------|
| CRITICAL | {finding} | {file:line} | {fixed/needs-action} |
| HIGH | {finding} | {file} | {fixed/needs-action} |
| MEDIUM | {finding} | {file} | {fixed/needs-action} |

### Files Created/Modified
- .env.example — Environment variable template ({N} variables)
- .gitignore — Added secret file patterns
- {provider config files}
- {application secret loader}

### Secrets Inventory
| Secret | Provider | Rotation | CI/CD |
|--------|----------|----------|-------|
| DATABASE_URL | {provider} | 30 days | configured |
| JWT_SECRET | {provider} | 90 days | configured |

### Immediate Actions Required
{ordered list of manual actions needed, starting with credential rotation}

============================================================ NEXT STEPS

1. Rotate any credentials that were exposed in source or git history
2. Set up the secrets provider account/project if not already done
3. Migrate existing .env values to the secrets provider
4. Update deployment scripts to pull secrets from provider at runtime
5. Enable GitHub secret scanning and push protection on the repository
6. Schedule quarterly secret rotation reviews

============================================================ SELF-EVOLUTION TELEMETRY

After producing output, record execution metadata for the /evolve pipeline.

Check if a project memory directory exists:

• Look for the project path in ~/.claude/projects/
• If found, append to skill-telemetry.md in that memory directory

Entry format:

### /secrets — {{YYYY-MM-DD}}
- Outcome: {{SUCCESS | PARTIAL | FAILED}}
- Self-healed: {{yes — what was healed | no}}
- Iterations used: {{N}} / {{N max}}
- Bottleneck: {{phase that struggled or "none"}}
- Suggestion: {{one-line improvement idea for /evolve, or "none"}}

Only log if the memory directory exists. Skip silently if not found. Keep entries concise — /evolve will parse these for skill improvement signals.

============================================================ DO NOT

• Do NOT print, log, or display actual secret values in any output
• Do NOT commit real secrets to any file, even temporarily
• Do NOT delete .env files without warning — they may be the only copy of secrets
• Do NOT run git filter-repo or git filter-branch without explicit user confirmation
• Do NOT disable git history scanning — secrets in history are a real risk
• Do NOT use symmetric encryption for secrets at rest — use provider-managed encryption
• Do NOT create IAM users with long-lived credentials — prefer roles and OIDC federation
• Do NOT store secrets in Terraform state without encrypting the state backend
• Do NOT skip the audit phase — always scan before configuring