Skip to content

docs(security-guide): add LLM Safe Haven to References (re: #1585) #2034

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
affaan-m merged 1 commit into from
May 25, 2026
+1 −0
Merged

docs(security-guide): add LLM Safe Haven to References (re: #1585) #2034

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 1 addition & 0 deletions the-security-guide.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -438,6 +438,7 @@ Scan your setup: [github.com/affaan-m/agentshield](https://github.com/affaan-m/a
- Microsoft Security, "AI Recommendation Poisoning" (February 10, 2026): [microsoft.com](https://www.microsoft.com/en-us/security/blog/2026/02/10/ai-recommendation-poisoning/)
- Snyk, "ToxicSkills: Malicious AI Agent Skills in the Wild": [snyk.io](https://snyk.io/blog/toxicskills-malicious-ai-agent-skills-clawhub/)
- Snyk `agent-scan`: [github.com/snyk/agent-scan](https://github.com/snyk/agent-scan)
- LLM Safe Haven (fail-closed runtime hooks, threat model, hardening guides for Claude Code/Cursor/Windsurf/Copilot/Codex/Aider/Cline): [github.com/pleasedodisturb/llm-safe-haven](https://github.com/pleasedodisturb/llm-safe-haven)
Copy link
Copy Markdown
Contributor

@coderabbitai coderabbitai Bot May 21, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

⚠️ Potential issue | 🟡 Minor | ⚡ Quick win

🧩 Analysis chain

🏁 Script executed:

#!/bin/bash
# Description: Verify the LLM Safe Haven repository exists and contains claimed features

# Check if repository is accessible
echo "=== Checking repository accessibility ==="
curl -s -o /dev/null -w "%{http_code}" https://github.com/pleasedodisturb/llm-safe-haven

# Clone and inspect repository structure
echo -e "\n=== Checking repository contents ==="
git clone --depth 1 https://github.com/pleasedodisturb/llm-safe-haven /tmp/llm-safe-haven 2>&1 || echo "Repository not found or not accessible"

if [ -d /tmp/llm-safe-haven ]; then
  echo -e "\n=== Repository structure ==="
  cd /tmp/llm-safe-haven
  ls -la
  
  echo -e "\n=== Checking for threat model documentation ==="
  fd -t f -e md . | xargs rg -l "threat model|OWASP" || echo "No threat model documentation found"
  
  echo -e "\n=== Checking for hardening guides ==="
  fd -t f -e md . | xargs rg -l -i "claude code|cursor|windsurf|copilot|codex|aider|cline" || echo "No hardening guides found"
  
  echo -e "\n=== Checking for runtime hooks ==="
  fd -t f | xargs rg -l "hook|runtime|fail-closed" || echo "No runtime hook implementation found"
  
  echo -e "\n=== Checking README for feature claims ==="
  cat README.md 2>/dev/null || echo "No README found"
fi

Repository: affaan-m/ECC

Length of output: 6332

🏁 Script executed:

#!/bin/bash
set -euo pipefail

REPO_DIR="/tmp/llm-safe-haven"

echo "== Searching for 'fail-closed' / runtime hook wording =="
rg -n -i "fail-closed|fail closed|runtime hook|failclosed" "$REPO_DIR" || true

echo
echo "== Checking threat model & hardening docs presence (quick index) =="
ls -1 "$REPO_DIR"/docs | sed -n '1,200p'
ls -1 "$REPO_DIR"/docs/hardening 2>/dev/null | sed -n '1,200p'
ls -1 "$REPO_DIR"/hooks | sed -n '1,200p'

echo
echo "== Checking for 'AgentShield' mention =="
rg -n -i "AgentShield" "$REPO_DIR" || true

Repository: affaan-m/ECC

Length of output: 990

🏁 Script executed:

#!/bin/bash
set -euo pipefail
REPO_DIR="/tmp/llm-safe-haven"

echo "== Search for Codex / Cline documentation =="
rg -n -i "codex|codex cli" "$REPO_DIR" || true
rg -n -i "cline" "$REPO_DIR" || true

echo
echo "== Search for OWASP Agentic Top 10 / threat model =="
rg -n -i "OWASP|agentic top 10|threat model" "$REPO_DIR"/docs "$REPO_DIR"/README.md "$REPO_DIR"/CLAUDE.md || true

echo
echo "== Check whether docs/hardening includes those agent pages =="
ls -1 "$REPO_DIR"/docs/hardening 2>/dev/null || true

Repository: affaan-m/ECC

Length of output: 21845

Line 441: Keep the entry, but tighten the scope of the feature claims

  • The GitHub link is reachable, and the repo includes fail-closed runtime hook behavior (e.g., hooks/secret-guard.js, hooks/bash-firewall.js) plus a threat model (docs/threat-model.md).
  • Hardening guides exist for Claude Code, Cursor, Windsurf, GitHub Copilot, and Aider; Codex CLI and Cline are supported/documented in the repo (e.g., README.md), but dedicated pages under docs/hardening/ for Codex/Cline weren’t found.
  • No “AgentShield” mention exists in the repo, so any “complements AgentShield” framing should be removed or supported with a specific reference.
  • Since this is the author’s own project, a quick maintainer quality/relevance check is still warranted before publishing.
🤖 Prompt for AI Agents 
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@the-security-guide.md` at line 441, Keep the "LLM Safe Haven" entry but
narrow the claims: update the line to state that the repo provides fail-closed
runtime hooks and a threat model (referencing hooks/secret-guard.js,
hooks/bash-firewall.js and docs/threat-model.md), list the hardening guides that
are present (Claude Code, Cursor, Windsurf, GitHub Copilot, Aider) and note that
Codex CLI and Cline are mentioned in README.md but do not have dedicated docs
under docs/hardening/, remove the phrase "complements AgentShield" (or add a
specific AgentShield reference if you can cite one), and add a brief
maintainer-quality/relevance check note before publishing.

- Hunt.io, "CVE-2026-25253 OpenClaw AI Agent Exposure" (February 3, 2026): [hunt.io](https://hunt.io/blog/cve-2026-25253-openclaw-ai-agent-exposure)
- OpenAI, "Designing AI agents to resist prompt injection" (March 11, 2026): [openai.com](https://openai.com/index/designing-agents-to-resist-prompt-injection/)
- OpenAI Codex docs, "Agent network access": [platform.openai.com](https://platform.openai.com/docs/codex/agent-network)
Expand Down