Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

Get Started Free →

A5c AI/Security Hardening — measured AI agent skill | DecimalAI

a5c-ai/security-hardening

SkillScore55Maintained

AIDefence security layer with prompt injection blocking, input validation, sandboxed execution, output sanitization, and STRIDE threat modeling.

by a5c-ai

Links this skill to your agents and keeps it current with the author’s updates.

Automatic

Security & AppSec pack

decimalai skills pull a5c-ai-security-hardening

Get this skill — curl · raw file

curl -fsSL https://app.decimal.ai/s/a5c-ai-security-hardening/SKILL.md

Raw file: /s/a5c-ai-security-hardening/SKILL.md ↗ · machine-readable summary at /s/a5c-ai-security-hardening.json ↗

In productionGlobal

how this skill behaves on real traffic · last 30 days

No production data for this skill yet — usage numbers appear here once agents start running it.

Anonymized aggregate across every org using this skill — proof from real production use, not a demo.

Does this skill help? · Benchmark: with vs. withoutsign in to run this on your agents

22 test cases from the publisher’s eval.yaml · each run twice on gemini-3.6-flash — with the skill injected, then without · last run 9d ago

Yes — +50% lift-33% tokens0% turns55% of whole cases pass with skill (12/22)

Test case	Without → With	Effect	Δ tokens	Δ turns
case-02	✗→✓	▲ Improved	12%	0%
case-05	✗→✓	▲ Improved	-20%	0%
case-06	✗→✓	▲ Improved	-50%	0%
case-07	✗→✓	▲ Improved	34%	0%
case-08	✗→✓	▲ Improved	-73%	0%

Showing 5 of 22 cases

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v4 — 7/31/2026

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

a5c-ai / babysitter

library/methodologies/ruflo/skills/security-hardening/SKILL.md

View on GitHub ↗

License MITImported 7/31/2026Version v1

Skill contents

SKILL.md

When processing untrusted inputs
Security audits of agent-generated code
Compliance verification (OWASP Top 10, CIS)

AIDefence Layers

Prompt Injection Detection - Pattern + heuristic blocking
Input Validation - Path traversal, type coercion, parameter sanitization
Static Analysis (SAST) - Vulnerability scanning, CWE matching
Sandboxed Execution - Network isolation, filesystem restrictions, resource limits
Output Sanitization - Secrets, PII, injection vector redaction

Security Levels

| Level | Layers | Use Case | |-------|--------|----------| | standard | SAST + validation + sanitization | Routine audits | | elevated | + threat modeling + compliance | Pre-release audits | | maximum | + sandbox + full STRIDE + remediation | Critical systems |

Agents Used

agents/security-auditor/ - Vulnerability detection
agents/reviewer/ - Code quality verification

Tool Use

Invoke via babysitter process: methodologies/ruflo/ruflo-security-audit

Want to see if a5c-ai/security-hardening would activate on your phrasing? Try it before you install — no LLM call, no signup.

Other measured skills in the registry, with their headline benchmark lift.