agent-deck
Safety scan
A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.
Check it deeper on your machine
It has changed since its scan on 11 Nov 2026.
As a Claude Code plugin, install it and then ask Claude Code:
/plugin marketplace add keithmackay/mackayi/plugin install scanskillsafety@mackayiDo a deep scan of https://github.com/asheshgoplani/agent-deck with scanskillsafety
Or from a terminal, clone it and run it:
git clone https://github.com/keithmackay/scanskillsafety-skill ~/.claude/skills/scanskillsafetynode ~/.claude/skills/scanskillsafety/cli.cjs --deep 'https://github.com/asheshgoplani/agent-deck'
It checks the exact commit: every file, its packages against OSV (api.osv.dev), how it compares with this page's scan (findsafeskills.com receives the repository's address), and, if you agree, Claude reading every file with your own Claude subscription. It never runs the code.
Badge for your README
Markdown
[](/listings/asheshgoplani--agent-deck)HTML
<a href="/listings/asheshgoplani--agent-deck"><img src="/api/listings/asheshgoplani--agent-deck/badge.svg" alt="findsafeskills safety rating" /></a>Install
Install instructions from the README
As written by the repo's author. findsafeskills has not verified these steps.
These steps run a script from the network that findsafeskills does not scan.
**Works on:** macOS, Linux, Windows (WSL)
curl -fsSL https://raw.githubusercontent.com/asheshgoplani/agent-deck/main/install.sh | bash
Then run: `agent-deck`
<details> <summary>Other install methods</summary>
**Homebrew**
brew install asheshgoplani/tap/agent-deck
**Go**
go install github.com/asheshgoplani/agent-deck/cmd/agent-deck@latest
**From Source**
git clone https://github.com/asheshgoplani/agent-deck.git && cd agent-deck && make install
</details>
<details> <summary>Uninstalling</summary>
agent-deck uninstall # Interactive uninstall agent-deck uninstall --keep-data # Remove binary only, keep sessions
See [Troubleshooting](skills/agent-deck/references/troubleshooting.md#uninstalling) for full details.
</details>
<details> <summary>Shell completion</summary>
`agent-deck completion <bash|zsh|fish>` prints a completion script for commands and subcommands (`agent-deck se<Tab>` → `session`, `agent-deck session st<Tab>` → `start`/`stop`), plus live resource names where it matters — `agent-deck remote update <Tab>` offers your configured remotes, `agent-deck session set-parent <Tab> <Tab>` offers session titles at both positions, and `agent-deck -p <Tab>` offers your profiles.
# bash echo 'source <(agent-deck completion bash)' >> ~/.bashrc # zsh echo 'source <(agent-deck completion zsh)' >> ~/.zshrc # fish agent-deck completion fish > ~/.config/fish/completions/agent-deck.fish
Open a new shell (or re-source the config file) afterwards.
</details>
What was scanned
The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.
.claude-plugin/marketplace.json · Manifest · 22 lines
View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)
This file has changed since it was scanned, so this is not exactly the text that was scanned, and no line is marked. The findings are listed with the safety scan above.
- {
- "name": "agent-deck",
- "owner": {
- "name": "Ashesh Goplani",
- "email": "ashesh.goplani96@gmail.com"
- },
- "metadata": {
- "description": "Skills for managing AI coding agent sessions with agent-deck CLI",
- "version": "1.0.0"
- },
- "plugins": [
- {
- "name": "agent-deck",
- "description": "Complete guide for managing AI coding agent sessions via agent-deck CLI - session lifecycle, MCP management, groups, profiles, and automation",
- "source": "./",
- "strict": false,
- "skills": [
- "./skills/agent-deck"
- ]
- }
- ]
- }
README.md · README · 106 lines
View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)
This file has changed since it was scanned, so this is not exactly the text that was scanned, and no line is marked. The findings are listed with the safety scan above.
- <div align="center">
- <!-- agent-deck mark: status dots on Calm Navy -->
- <img src="site/logo.svg" alt="Agent Deck Logo" width="96">
- # Agent Deck
- [](https://github.com/asheshgoplani/agent-deck/stargazers)
- [](https://github.com/asheshgoplani/agent-deck/releases)
- [](https://go.dev)
- [](LICENSE)
- [](https://github.com/asheshgoplani/agent-deck)
- [](https://github.com/asheshgoplani/agent-deck/releases)
- [](https://discord.gg/e4xSs6NBN8)
- [Install](#installation) . [Quick Start](#quick-start) . [Connect your assistant](#connect-your-assistant) . [Docs](#documentation) . [Discord](https://discord.gg/e4xSs6NBN8) . [FAQ](docs/guide/faq.md)
- </div>
- Agent Deck runs all your AI coding agents (Claude Code, Codex, Gemini CLI, OpenCode and more) in one terminal and shows which ones are working, waiting, or done.
- https://github.com/user-attachments/assets/e4f55917-435c-45ba-92cc-89737d0d1401
- ## Installation
- Works on macOS, Linux and Windows (WSL).
- ```bash
- curl -fsSL https://raw.githubusercontent.com/asheshgoplani/agent-deck/main/install.sh | bash
- ```
- Or with Homebrew:
- ```bash
- brew install asheshgoplani/tap/agent-deck
- ```
- Other ways to install, updates, uninstall and shell completion: [Install and update](docs/guide/install.md).
- ## Quick Start
- ```bash
- cd ~/code/my-project # any project folder
- agent-deck launch . -c claude # create and start a Claude Code session here
- agent-deck # open the deck, pick the session, press Enter
- ```
- The keys you need first:
- | Key | Action |
- |-----|--------|
- | `Enter` | Attach to session |
- | `Ctrl+Q` | Detach from session |
- | `n` | New session |
- | `/` | Search |
- | `Alt+W` | Show me around: what you can do, your agents and sign-in |
- | `Alt+N` | What's new: this version's highlights and your reports it fixed |
- | `?` | Full help |
- More commands and every key: [Quick start and keys](docs/guide/quick-start.md).
- ## Connect your assistant
- Paste this into your assistant (Hermes, OpenClaw, Grok, Claude, ChatGPT, Codex...):
- ```text
- Read https://raw.githubusercontent.com/asheshgoplani/agent-deck/main/docs/assistants/SETUP-PROMPT.md and follow it to connect yourself to agent-deck on my machine. Ask me before each step that changes anything.
- ```
- It checks agent-deck, connects itself, saves its own config and tests it. Or run `agent-deck connect` and paste what it prints, or use the Connect page in `agent-deck web`.
- Your assistant can only read your sessions at first (read-only). You choose if it may do more. Details: [Connect an assistant](https://github.com/asheshgoplani/agent-deck/blob/main/docs/assistants/README.md).
- ## Documentation
- - [Guide index](docs/guide/README.md): every guide and reference in one list
- - [Install and update](docs/guide/install.md)
- - [Quick start and keys](docs/guide/quick-start.md)
- - [Sessions, groups and config](docs/guide/sessions.md): MCP and skills managers, groups, accounts, search, status, forking, archive, feedback
- - [Worktrees and sandboxes](docs/guide/worktrees.md): git worktrees, Docker sandbox
- - [Conductor and watchers](docs/guide/conductor.md): run a fleet from your phone
- - [Tools, costs and quotas](docs/guide/tools.md): supported agents, cost dashboard, `agent-deck usage`
- - [Remote machines and the web UI](docs/guide/remote-and-web.md)
- - [FAQ and asking an AI](docs/guide/faq.md)
- - [CLI reference](skills/agent-deck/references/cli-reference.md) . [Config reference](skills/agent-deck/references/config-reference.md) . [Troubleshooting](skills/agent-deck/references/troubleshooting.md)
- - [Contributing and development](docs/guide/contributing.md) . [CONTRIBUTING.md](CONTRIBUTING.md) . [CHANGELOG.md](CHANGELOG.md)
- ## Star History
- If Agent Deck saves you time, give us a star! It helps others discover the project.
- [](https://star-history.com/#asheshgoplani/agent-deck&Date)
- ## License
- MIT License — see [LICENSE](LICENSE)
- ---
- <div align="center">
- Built with [Bubble Tea](https://github.com/charmbracelet/bubbletea) and [tmux](https://github.com/tmux/tmux)
- **[Docs](skills/agent-deck/references/) . [Discord](https://discord.gg/e4xSs6NBN8) . [Issues](https://github.com/asheshgoplani/agent-deck/issues) . [Discussions](https://github.com/asheshgoplani/agent-deck/discussions)**
- </div>
skills/agent-deck/scripts/self-improvement/sanitize.py · Bundled script · 311 lines · 1 finding marked
- #!/usr/bin/env python3
- """Sanitize a FINDINGS.md / CAPABILITIES.md for public sharing.
- Hardened version — privacy-first. Aggressive redaction over signal preservation
- on disagreement. Always emits a parallel mapping JSON locally for traceback.
- Redacts:
- - Absolute home paths (/home/<user>, /Users/<user>)
- - Scratch / log paths (/tmp/..., /var/...)
- - Worktree names (.worktrees/<name>)
- - Session UUID prefixes ([abc12345] -> [event-NNN])
- - Report-filename SIDs (5ed9163e.md -> report-NN.md)
- - API tokens / secrets (sk-..., gh[ps]_..., AKIA..., JWTs, Telegram tokens, generic TOKEN=)
- - Email addresses with TLD (user@example.com)
- - IP addresses (v4) (any A.B.C.D)
- - Long hex strings (commit SHAs) (>=12 hex chars)
- - User-supplied substitutions (innotrade, ryan, internal domains, etc.)
- - Maintainer GitHub handle (@asheshgoplani -> <maintainer>)
- - Generic GitHub repo refs (foo/bar -> <org>/bar if foo is in --map)
- Audits afterwards and flags any line still containing suspicious patterns
- for HUMAN REVIEW before sharing.
- Usage:
- python sanitize.py <input.md> <output.md> [--map key=value ...]
- The mapping JSON written alongside the output is for LOCAL traceback only.
- Never share it.
- """
- from __future__ import annotations
- import argparse
- import json
- import re
- import sys
- from pathlib import Path
- # -- Paths -------------------------------------------------------------------
- HOME_PATH_RE = re.compile(r"/(home|Users)/[a-zA-Z0-9_.-]+")
- TMP_PATH_RE = re.compile(r"/tmp/[a-zA-Z0-9._/-]+")
- VAR_PATH_RE = re.compile(r"/var/[a-zA-Z0-9._/-]+")
- WORKTREE_RE = re.compile(r"\.worktrees/[a-zA-Z0-9._/-]+")
- # -- Identifiers in our findings docs ----------------------------------------
- UUID_PREFIX_RE = re.compile(r"\[([0-9a-f]{8})\]") # citation tokens [346296cb]
- REPORT_SID_RE = re.compile(r"\b([0-9a-f]{8})\.md\b") # report-file references
- FULL_UUID_RE = re.compile(r"\b[0-9a-f]{8}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{12}\b")
- LONG_HEX_RE = re.compile(r"\b[a-f0-9]{12,}\b") # commit SHAs / content hashes
- # -- Secrets / tokens (HIGH PRIORITY) ----------------------------------------
- ANTHROPIC_KEY_RE = re.compile(r"sk-(?:ant-)?[a-zA-Z0-9_-]{20,}")
- OPENAI_KEY_RE = re.compile(r"sk-(?:proj-)?[a-zA-Z0-9_-]{40,}")
- GITHUB_TOKEN_RE = re.compile(r"gh[opsu]_[a-zA-Z0-9]{20,}")
- AWS_KEY_RE = re.compile(r"AKIA[0-9A-Z]{16}")
- TELEGRAM_BOT_RE = re.compile(r"\b\d{8,12}:[A-Za-z0-9_-]{30,}\b")
- JWT_RE = re.compile(r"\beyJ[a-zA-Z0-9_-]+\.[a-zA-Z0-9_-]+\.[a-zA-Z0-9_-]+\b")
- GENERIC_SECRET_RE = re.compile(
- r"(?i)\b(token|key|secret|password|api[_-]?key|auth(?:orization)?)\s*[=:]\s*['\"]?([a-zA-Z0-9_+/=.-]{12,})['\"]?"
- )
- SECRET_PATTERNS = [
- ("anthropic_key", ANTHROPIC_KEY_RE),
- ("openai_key", OPENAI_KEY_RE),
- ("github_token", GITHUB_TOKEN_RE),
- ("aws_access_key", AWS_KEY_RE),
- ("telegram_bot_token", TELEGRAM_BOT_RE),
- ("jwt", JWT_RE),
- ]
- # -- Other PII ---------------------------------------------------------------
- EMAIL_RE = re.compile(r"\b[a-zA-Z0-9._%+-]+@[a-zA-Z0-9.-]+\.[a-zA-Z]{2,}\b")
- IPV4_RE = re.compile(r"\b(?:(?:25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9]?)\.){3}(?:25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9]?)\b")
- GITHUB_HANDLE_RE = re.compile(r"@asheshgoplani\b")
- REPO_OWNER_RE = re.compile(r"\basheshgoplani/")
- # -- Audit (post-sanitize check) ---------------------------------------------
- # After sanitization, flag any line still containing these patterns.
- SUSPICIOUS_PATTERNS = [
- ("possible_long_hex", re.compile(r"\b[a-f0-9]{20,}\b")),
- ("possible_b64_blob", re.compile(r"\b[A-Za-z0-9+/=]{40,}\b")),
- ("env_var_with_value", re.compile(r"\b[A-Z_][A-Z0-9_]{4,}\s*[=:]\s*[^\s,]{8,}")),
- ("possible_phone", re.compile(r"\+\d[\d\s\-]{7,}")),
- ("private_ip_range", re.compile(r"\b(10|172|192\.168|100)\.\d")), # rough heuristic
- ]
- def redact_with_label(label: str, text: str) -> str:
- return f"<REDACTED:{label}>"
- def sanitize(text: str, user_subs: dict[str, str]) -> tuple[str, dict]:
- """Return (sanitized_text, mapping)."""
- mapping: dict = {
- "user_subs": user_subs,
- "home_paths_seen": [],
- "tmp_paths_seen": [],
- "var_paths_seen": [],
- "worktree_paths_seen": [],
- "uuid_prefixes": {},
- "report_sids": {},
- "full_uuids_seen": 0,
- "long_hex_seen": 0,
- "secrets_redacted_by_kind": {},
- "emails_seen": 0,
- "ipv4_seen": 0,
- }
- # 1. User-supplied substitutions first (business names, conductor names)
- for raw, sub in user_subs.items():
- if raw in text:
- text = text.replace(raw, sub)
- # 2. Secrets — HIGHEST PRIORITY, do these first before any other pattern
- for kind, pat in SECRET_PATTERNS:
- def make_repl(k):
- def repl(m):
- mapping["secrets_redacted_by_kind"][k] = (
- mapping["secrets_redacted_by_kind"].get(k, 0) + 1
- )
- return redact_with_label(k, m.group(0))
- return repl
- text = pat.sub(make_repl(kind), text)
- # Generic KEY=VALUE secrets — only redact the value
- def repl_generic_secret(m):
- kind = "generic_secret"
- mapping["secrets_redacted_by_kind"][kind] = (
- mapping["secrets_redacted_by_kind"].get(kind, 0) + 1
- )
- return f"{m.group(1)}=<REDACTED:{kind}>"
- text = GENERIC_SECRET_RE.sub(repl_generic_secret, text)
- # 3. Emails
- def repl_email(m):
- mapping["emails_seen"] += 1
- return "<REDACTED:email>"
- text = EMAIL_RE.sub(repl_email, text)
- # 4. IPv4 addresses
- def repl_ip(m):
- mapping["ipv4_seen"] += 1
- return "<REDACTED:ipv4>"
- text = IPV4_RE.sub(repl_ip, text)
- # 5. GitHub handle and repo owner
- text = GITHUB_HANDLE_RE.sub("<maintainer>", text)
- text = REPO_OWNER_RE.sub("<owner>/", text)
- # 6. Paths
- seen_home = set()
- seen_tmp = set()
- seen_var = set()
- seen_wt = set()
- def repl_home(m):
- s = m.group(0)
- if s not in seen_home:
- seen_home.add(s)
- mapping["home_paths_seen"].append(s)
- return f"/{m.group(1)}/<user>"
- text = HOME_PATH_RE.sub(repl_home, text)
- def repl_tmp(m):
- s = m.group(0)
- if s not in seen_tmp:
- seen_tmp.add(s)
- mapping["tmp_paths_seen"].append(s)
- return "/tmp/<scratch>"
- text = TMP_PATH_RE.sub(repl_tmp, text)
- def repl_var(m):
- s = m.group(0)
- if s not in seen_var:
- seen_var.add(s)
- mapping["var_paths_seen"].append(s)
- return "/var/<system>"
- Note: Usual in code (tag names in paths and patterns): Contains a <system> block, a common wrapper for instructions hidden in a tool description
- Line 177 on GitHub
- text = VAR_PATH_RE.sub(repl_var, text)
- def repl_wt(m):
- s = m.group(0)
- if s not in seen_wt:
- seen_wt.add(s)
- mapping["worktree_paths_seen"].append(s)
- return ".worktrees/<branch>"
- text = WORKTREE_RE.sub(repl_wt, text)
- # 7. Full session UUIDs (before short ones)
- def repl_full_uuid(m):
- mapping["full_uuids_seen"] += 1
- return "<REDACTED:session-uuid>"
- text = FULL_UUID_RE.sub(repl_full_uuid, text)
- # 8. UUID prefixes -> deterministic [event-NNN] tokens
- uuid_order: dict[str, str] = {}
- def repl_uuid_prefix(m):
- raw = m.group(1)
- if raw not in uuid_order:
- uuid_order[raw] = f"event-{len(uuid_order) + 1:03d}"
- return f"[{uuid_order[raw]}]"
- text = UUID_PREFIX_RE.sub(repl_uuid_prefix, text)
- mapping["uuid_prefixes"] = uuid_order
- # 9. Report SIDs -> deterministic report-NN.md
- sid_order: dict[str, str] = {}
- def repl_sid(m):
- raw = m.group(1)
- if raw not in sid_order:
- sid_order[raw] = f"report-{len(sid_order) + 1:02d}"
- return f"{sid_order[raw]}.md"
- text = REPORT_SID_RE.sub(repl_sid, text)
- mapping["report_sids"] = sid_order
- # 10. Long hex (commits, hashes) — replace with placeholder
- def repl_long_hex(m):
- mapping["long_hex_seen"] += 1
- return "<hash>"
- text = LONG_HEX_RE.sub(repl_long_hex, text)
- return text, mapping
- def audit(text: str) -> list[dict]:
- """Run post-sanitize audit. Returns list of suspicious finds (line + reason)."""
- findings = []
- for ln_no, line in enumerate(text.splitlines(), 1):
- # Skip lines that are clearly our own REDACTED markers / generated tokens
- for label, pat in SUSPICIOUS_PATTERNS:
- for m in pat.finditer(line):
- # Filter false positives: our REDACTED labels and event tokens
- snippet = m.group(0)
- if snippet.startswith("<REDACTED") or snippet.startswith("event-") or snippet.startswith("report-"):
- continue
- findings.append({
- "line": ln_no,
- "label": label,
- "snippet": snippet[:120],
- "context": line.strip()[:200],
- })
- break # only first finding per line per pattern, to avoid spam
- return findings
- def main() -> int:
- ap = argparse.ArgumentParser()
- ap.add_argument("input", type=Path)
- ap.add_argument("output", type=Path)
- ap.add_argument(
- "--map",
- nargs="*",
- default=[],
- help="User substitutions: key=value (e.g. innotrade=<work-conductor>)",
- )
- ap.add_argument("--no-audit", action="store_true", help="Skip post-audit (not recommended)")
- args = ap.parse_args()
- if not args.input.exists():
- print(f"Input not found: {args.input}", file=sys.stderr)
- return 1
- user_subs: dict[str, str] = {}
- for kv in args.map:
- if "=" not in kv:
- print(f"Bad --map entry: {kv} (expected key=value)", file=sys.stderr)
- return 2
- k, v = kv.split("=", 1)
- user_subs[k] = v
- text = args.input.read_text(encoding="utf-8")
- sanitized, mapping = sanitize(text, user_subs)
- args.output.write_text(sanitized, encoding="utf-8")
- map_path = args.output.with_suffix(args.output.suffix + ".map.json")
- map_path.write_text(json.dumps(mapping, indent=2), encoding="utf-8")
- audit_findings = [] if args.no_audit else audit(sanitized)
- audit_path = args.output.with_suffix(args.output.suffix + ".audit.json")
- audit_path.write_text(json.dumps(audit_findings, indent=2), encoding="utf-8")
- summary = {
- "input": str(args.input),
- "output": str(args.output),
- "mapping_path (LOCAL ONLY)": str(map_path),
- "audit_path": str(audit_path),
- "stats": {
- "in_bytes": len(text),
- "out_bytes": len(sanitized),
- "secrets_redacted_by_kind": mapping["secrets_redacted_by_kind"],
- "home_paths": len(mapping["home_paths_seen"]),
- "tmp_paths": len(mapping["tmp_paths_seen"]),
- "var_paths": len(mapping["var_paths_seen"]),
- "worktree_paths": len(mapping["worktree_paths_seen"]),
- "uuid_prefixes": len(mapping["uuid_prefixes"]),
- "report_sids": len(mapping["report_sids"]),
- "full_uuids": mapping["full_uuids_seen"],
- "long_hex": mapping["long_hex_seen"],
- "emails": mapping["emails_seen"],
- "ipv4": mapping["ipv4_seen"],
- "user_subs_applied": len(user_subs),
- },
- "audit": {
- "suspicious_findings": len(audit_findings),
- "must_review_before_sharing": len(audit_findings) > 0,
- },
- }
- print(json.dumps(summary, indent=2))
- return 0
- if __name__ == "__main__":
- sys.exit(main())
Ratings and comments
Sign in to rate this skill.
Sign inFree to use. If it’s helped, pass it on.