Updated daily · last run Nov 20, 07:02 UTC
Trending in the bucket
Every AI dev tool the internet posted about, ranked by who's talking about it. Sorted. Mostly.
- In the bucket
- 3.1K
- Posts read
- 42.7K
- Tossed out
- 33.7K
33 tools, ranked by sources, chatter and star growth. How ranking works
- 01
#1
Desktop app that finds and revokes permissions, MCP servers, trusted folders and secrets granted to AI coding agents, with live alerts and h
HN#claude-code#codex#gemini-cli1 post · 6 days ago★ 15 +5/wk
Stars:15
+5 this week
- 02
#2
Static security scanner with 91 rules, an MCP server and agent skill that audits AI-written code for security flaws.
GHHN#claude-code#codex#cursor2 posts · 12 days ago★ 74 +46/wk
Stars:74
+46 this week
- 03
#3
Security scanner for AI agent skills that detects vulnerabilities, prompt injection, data exfiltration, and supply-chain risks before instal
GH#claude-code#codex#gemini-cli1 post · 24 days ago★ 27K +1.2K/wk
Stars:27K
+1.2K this week
- 04
#4
Zero-setup, zero-latency sandbox for running AI coding agents with least-privilege filesystem, network, and credential policies.
HN#claude-code#codex#opencode3 posts · 22 days ago★ 5.1K +104/wk
Stars:5.1K
+104 this week
- 05
#5
Agent skill for WordPress security operations: audit, hardening, incident response, forensics and recovery, with or without BotBlocker.
GH#claude-code#opencode#codex1 post · 13 days ago★ 17
Stars:17
- 06
#6
Open-source daemon that discovers AI coding tools, inventories MCP servers and skills, applies sandbox policy, and connects devices to an LL
GHHN#claude-code#codex#cursor2 posts · 17 days ago★ 184 +12/wk
Stars:184
+12 this week
- 07
#7
Hook-based guardrail tool that scores every coding agent edit against AGENTS.md/CLAUDE.md project rules and forces same-turn repairs.
GH#claude-code#codex#opencode1 post · Sep 18★ 631 +3/wk
Stars:631
+3 this week
- 08#8
Open-source gate that scores coding agent Bash commands against a repo's GUARDRAILS.md and blocks destructive, secret-leaking, or policy-vio
- 09
#9
CLI agent harness that orchestrates coding agents for pentest, code audit, and incident-response engagements with a Kali runtime and durable
GHHN#claude-code#codex#opencode3 posts · Sep 22★ 34
Stars:34
- 10
#10
Uber's open-source security system for discovering, observing, and detecting threats from AI coding agents on enterprise endpoints.
HN#claude-code#cursor#codex1 post · Aug 6★ 2K
Stars:2K
- 11
#11
Repo of guardrails and 35 security Agent Skills plus a toolchain installer that turns AI coding agents into authorized security assistants.
GH#claude-code#cursor#codex1 post · Aug 8★ 11
Stars:11
- 12
#12
CLI that lints Agent Skills specs, audits unsafe instructions, and checks SKILL.md compatibility across coding agents.
GH#claude-code#cursor#codex1 post · Aug 13★ 12
Stars:12
- 13
#13
Security scanner that pairs static taint analysis with your coding agent to dynamically verify or refute vulnerabilities in local apps.
GH#claude-code#codex#opencode1 post · Aug 19★ 137
Stars:137
- 14
#14
Local encrypted secrets vault for coding agents; agents reference secrets by name and values are injected and masked, never shown to the mod
HN#claude-code#codex#opencode1 post · Aug 29★ 86
Stars:86
- 15
#15
CLI that stores signed per-commit evidence records of agent-written code: line attribution, agent intent, and test results.
GHHN#claude-code#codex#opencode2 posts · Sep 13★ 23
Stars:23
- 16
#16
Runtime permission firewall for AI agents: YAML policies, popup/Telegram approvals, and tamper-evident signed audit logs.
GH#claude-code#opencode#any-agent1 post · Sep 22★ 10
Stars:10
- 17
#17
CLI that discovers and scans installed agents, MCP servers, and skills for prompt injections, malware, and other security risks.
HN#claude-code#cursor#codex1 post · Aug 30★ 3.1K
Stars:3.1K
- 18
#18
An unofficial OpenCode plugin that reviews every 'ask' permission request with a tool-free AI model against a tenant policy, allowing, denyi
GH#opencode#plugin#any-model1 post · Aug 4★ 16
Stars:16
- 19
#19
Plugin collection that auto-approves harmless shell permission requests in coding agents using probabilistic AI judgments.
GH#claude-code#opencode#plugin1 post · Sep 18★ 30
Stars:30
- 20
#20
Local-only secret manager (SOPS+age) with an MCP server so AI coding agents can inject secrets into commands without ever seeing the values.
GH#mcp-server#claude-code#codex1 post · Aug 17★ 15
Stars:15
- 21
#21
A secrets vault that injects passwords and API keys into agent-run commands without exposing them, with guard hooks and skills for coding ag
GH#claude-code#cursor#codex1 post · Jul 21★ 71
Stars:71
- 22
#22
Guardrail kit for coding agents: deterministic checks, a fresh-context reviewer that commits fixes, and one-way-door PR triage.
GH#claude-code#codex#opencode1 post · Sep 26★ 15
Stars:15
- 23
#23
Kernel-layer runtime security for AI agents: per-agent file, exec, install and network policy enforcement with refusal audit logs.
GH#claude-code#codex#cursor1 post · Sep 21★ 11
Stars:11
- 24
#24
Security hook for coding agents that risk-scores tool calls, flags prompt injection in results, and checks skills and instruction files.
GH#claude-code#codex#cursor1 post · Sep 17★ 153
Stars:153
- 25#25
Local protection layer that replaces secrets with opaque handles in AI coding client requests, restoring real values only before local tools
HN#claude-code#codex#opencode1 post · Aug 5