172 packages matching “toolcall”
pi-thinking-only-guard
v0.1.0 · 2 months ago
Auto-recover trapped tool calls from thinking blocks for Qwen3.6 and similar models
No known vulnerabilities
@formloom/llm
v0.4.0 · 3 months ago
LLM integration layer for Formloom - prompt fragments, tool definitions, parser, and return-path helper for OpenAI, Anthropic, Gemini, Mistral, and Ollama
No known vulnerabilities
pi-tool-repair
v0.2.1 · 22 hours ago
Validate-then-repair extension for pi — fixes common LLM tool-call mistakes (null fields, stringified arrays, wrong field names, anchor bleed) before tools execute
No known vulnerabilities
@cursor-ai/january
v0.2.1 · 8 months ago
Experimental project
No known vulnerabilities
@benvargas/pi-claude-code-use
v2.2.0 · 5 days ago
Patch Anthropic OAuth payloads and auto-alias extension tools for Claude Code-style subscription use
No known vulnerabilities
pi-dynamic-context-pruning
v0.1.6 · 8 days ago
A pi extension that dynamically prunes stale context to keep sessions lean.
No known vulnerabilities
@bazilion/api-types
v0.12.2 · 8 days ago
Hermetic wire-shape types for the Bazilion HTTP/IPC surface (zero deps).
No known vulnerabilities
@sockt/types
v0.2.2 · 23 days ago
Shared TypeScript types, Zod schemas, and store/client interfaces for Sockt. Every other `@sockt/*` package depends on this one — it's the root of the dependency graph and defines every contract crossed between packages (task shapes, LLM messages, memory
No known vulnerabilities
@brightsy/types
v1.0.0 · 4 months ago
Shared types for Brightsy agents, chat, messages, tools, and plans
No known vulnerabilities
@d3ara1n/pi-editor-shell
v0.8.1 · 1 day ago
Replaces pi's default editor and status bar with a unified rounded-corner shell embedding status info in the border
No known vulnerabilities
@dudousxd/nestjs-agent-testing
v0.9.0 · 17 days ago
nestjs-agent testing — in-memory store/sink/quota + deterministic fake model
No known vulnerabilities
@antipopp/agno-types
v0.14.0 · 1 month ago
TypeScript types for Agno client libraries
No known vulnerabilities
@gurdy/sdk
v0.1.0 · 21 hours ago
Provenance enrichment for governed agent traffic — the on-ramp to gurdy-proxy.
No known vulnerabilities
@johnhenry/acpq
v0.0.1 · 18 days ago
Reactive session/turn store + permission broker for the Agent Client Protocol — embeddable ACP client state for web UIs, notebooks, and editors. On the official @agentclientprotocol/sdk.
No known vulnerabilities
npcts
v0.1.25 · 1 month ago
NPC TypeScript SDK - A comprehensive UI component library for building AI-powered applications
No known vulnerabilities
opencode-claude-code-plugin
v0.1.2 · 5 months ago
Claude Code CLI provider plugin for opencode
No known vulnerabilities
@danypops/pi-process-harness
v0.1.3 · 4 days ago
Spawns real `pi` processes (and companion daemons) for integration tests that need a real AgentSession genuinely deciding to call a tool -- scripted via Pi's own first-party faux model provider, no live LLM call. The sibling of @danypops/pi-extension-harn
No known vulnerabilities
@cogitator-ai/types
v0.22.3 · 20 days ago
Shared TypeScript types for Cogitator
No known vulnerabilities
gpt-toolkit
v2.4.11 · 2 years ago
Pass in some Zod schemas, get a statically typed AI agent back!
No known vulnerabilities
@paraph/sdk
v0.2.1 · 13 days ago
Paraph SDK — tamper-evident records and human-approval gates for AI agents. EU AI Act Articles 12 + 14.
No known vulnerabilities