prompt
@code-fixer-23/pi-prompt-guard
Pi package scaffold for pi-prompt-guard
+1
Packages
6.2k listings for pi
🛡 Guards & policy Block or rewrite dangerous tool calls, sandbox commands, enforce permissions.
323 results · page 4 of 7
extension
@shelken/pi-guard
Hard-block dangerous agent bash commands and secret path access
skill
@bacnh85/pi-ux
Anti-slop UI/UX design discipline for your Pi agent — anchors a lintable DESIGN.md, runs deterministic slop-audit gates (APCA contrast + tokens + states + slop tells), works with text-only models, ships reference design-system presets.
+1
extension
@xaccefy/pi-casefile
Offensive security case tracker for Pi Agent — bug bounties, CTFs, security audits
+2
extension
@false00/pi-elasticsearch
Production-focused Elasticsearch automation tools for the Pi coding agent — curated tools for search, documents, indices, cluster administration, security, lifecycle, and full API reach via the official JS client and raw REST access
extension
@johanthoren/jeff
Jeff is a model-native quality control plane and cooperative workflow protocol for trusted operators and friendly agents, not a security sandbox.
+1
extension
@itc-steve/pi-ask-complete
Pi extension: bottom-panel ask_user + bash permission gate (Allow / Allow permanently / Deny with reason) writing ~/.pi/agent/permission.json
+2
extension
pi-obsidian-vault
Agent-safe Obsidian vault access for Pi: auto-detect, retrieve, validate, plan, write, edit, manage, and explicitly destroy Markdown with human approval.
extension
pi-code-tool
Code-mode tool for the pi coding agent: a sandboxed Python meta-tool (via @pydantic/monty) that lets the agent write code against host tools and build ephemeral code tools
extension
pi-defender
Defense-in-depth protection for Pi coding agent. Blocks dangerous commands and protects sensitive files — rename of pi-damage-control.
+4
extension
@jerryan/pi-sanity
Pi extension for sanity checks on agent operations
extension
pi-extension-e2b
E2B cloud sandbox integration for pi — redirects all tool execution to a remote E2B sandbox
+4
extension
pi-minimal-permission-system
Minimal permission enforcement extension for the Pi coding agent.
extension
@arvoretech/pi-secret-firewall
Redacts secrets (env vars, .env files, token patterns) from the model context and tool outputs, exposing them only as $SECRET_* shell env vars the model can reference but never read
skillextension
agent-booster-pack
Agent Booster Pack for Pi: skills + Interface Design Gate runtime + proof-first runtime + whiteboarding guard + final value guard, in one install. Meta-package; depends on the four siblings.
+2
extension
pi-bash-confirm
Pi package for confirming bash commands before execution with Telegram notifications
+1
toolextension
@bacnh85/pi-windows-tools
Pi extension for Windows-native tool manipulation — shell profiles, path conversion, command execution, WSL bridge, safety policy, and developer tool discovery.
+1
extension
@xaccefy/pi-exploitsearch
ExploitSearch extension for Pi Agent — query preview.is security corpus for attack techniques, primitives, and bypasses
themeskill
zmarketplace
Cross-agent marketplace search: find, audit, and install plugins/skills/themes/prompts across pi, omp, claude code, opencode, gemini cli, and codex
+1
extension
pi-secret-guard
A pi extension that guards against committing secrets, API keys, and credentials to git repositories using hybrid regex + LLM review.
skill
@firstpick/pi-package-skill-lifecycle
Self-contained Pi package for skill lifecycle management: memory, skill-bank audit, evaluation, creation, and refinement tools.
+1
toolextension
pi-opa-net
OPA-backed bash command guard for the pi ecosystem — structured decision-output.v1 JSON, fail-open default, Claude Code hook protocol compatible. Agent-agnostic engine + CLI.
extension
@eggmasonvalue/pi-subagent
Observable, steerable, isolated-context subagents for pi — resume any child mid-task, flush partial results on interrupt/timeout, lean machine-facing envelope, optional model allowlist.
skill
@firstpick/pi-skill-deep-research
Agents should invoke this skill for high-stakes or complex research needing multi-source evidence, scientific/technical fact-checking, decision traces, or rigorous verification. Runs deterministic two-phase research with schema/policy validation.
extension
pi-enclave
VM-isolated sandbox for pi with automatic secret protection · from yapp
+2
extension
pi-vscode-sr
Code diff assistant for VS Code.
promptextension
pi-secure-it
In-process tool guard and subagent posture hardening for Pi coding agent. Enforces filesystem allow/deny lists and domain allowlists on read, write, edit, fetch_content, web_search, and get_search_content — with interactive ask-tier prompts to grant acces
+1
extension
@sysid/pi-sandbox
OS + app sandboxing extension for the pi coding agent
+1
skill
@firstpick/pi-skill-subagent-governance
Portable Agent Skill that governs delegation admissibility for a parent orchestrator, covering zero-or-multiple fanout, the static two-worker minimum, role-fit balance, one-writer isolation, worker handoffs, bounded retry safety, and reviewer finding disp
extension
@chrischow/pi-lockdown
A Pi extension to add security constraints to agent tool usage.
prompt
sterlai
Terminal AI agent for Stellar. Chat-driven payments, path payments, trustlines, and friendbot funding on testnet, with a separate signer daemon that prompts for explicit approval on every signature.
+3
extension
pi-approval-guardian
A fail-closed-by-default Pi approval gate that reviews shell actions, private data access, and sensitive file mutations with an isolated risk model and explicit temporary bypass.
+1
toolextension
pi-sandbox-proxy
pi coding-agent extension that intercepts network operations with approval flows, vulnerability scanning, and supply chain security enforcement.
+1
extension
@ramarivera/pi-model-guardrails
Deterministic destructive-command guard + inviolable-constraint policy engine and deviation state machine for the Pi coding agent
+2
extension
@guygrigsby/pi-subagent-routing
Inject a subagent model-routing policy: the orchestrator plans and judges on its own live model and fans implementation out to cheaper subagents, reviewing before it accepts.
extension
@mammothb/pi-permissions
Permission enforcement extension for pi — gates tool calls, bash commands (via arbiter executable), and file paths
+1
extension
@0xkobold/pi-gateway
Hermes-style messaging gateway for pi - multi-platform agent with sessions, security, and background tasks
+2
extension
@victor-software-house/pi-credential-vault
Managed-provider credential vault for Pi with built-in age, keychain, and passthrough backends
extension
pi-ward
File access guard for the pi coding agent — declarative filesystem boundaries.
extension
pi-permission-gate
Config-driven permission system for Pi agents. Deny-by-default tool restriction with glob matching, path normalization, self-protection, and structured logging.
extension
pi-hooks
Collection of pi extensions (checkpoint, lsp, permission, ralph-loop, repeat)
+3
promptextension
pi-command-guard
A pi extension that intercepts potentially dangerous bash commands and prompts for confirmation before execution
skill
@firstpick/pi-skill-research-orchestration
Agents should invoke this skill for broad multi-claim research projects needing planning, parallel investigation, source merging, gap closure, citation audit, and final synthesis when narrower research skills are insufficient.
extension
pi-container-sandbox
pi coding-agent extension that runs every read/write/edit/bash op inside a per-session Linux container (Apple `container` or Docker).
+2
extension
@oddsjam/pi-sandbox
OS-level sandbox for pi using @anthropic-ai/sandbox-runtime, with an in-pi configure wizard, shift+tab toggle, and longest-prefix project configs.
+1
skill
@yammd/forge-skills
Global AI Skills Ecosystem for code review and audit
promptextension
pi-claude-sandbox
Claude-style OS-level sandboxing for pi with interactive permission prompts. Coexists with pi-tool-display and other bash-overriding extensions. Forked from carderne/pi-sandbox.
extension
@smallbatchcode/pi-slash-command-guard
Pi extension that blocks mistyped or unknown slash commands from being sent as messages.