@ogulcancelik/pi-model-agents
Load model-specific AGENTS.md instructions in pi
6.2k listings for pi
58 results · page 1 of 2
Load model-specific AGENTS.md instructions in pi
A Pi extension that observes coding sessions and distills patterns into reusable instincts.
Append model-specific content to Pi system prompt before Project Context
Reusable Pi prompt templates for maintaining durable agent memory.
File-based troubleshooting LEARNINGS archive for Pi agents, with retrieval workflow, summary prompts, and sync scripts.
Pi extension: inspect the system prompt and full LLM context in scrollable overlay popups
PI extension that injects confirmed personal facts from the Bee wearable assistant into the system prompt for personalized responses
pi extension for loading extra context files into the system prompt.
Pi coding agent extension that wires codegraph (tree-sitter knowledge graph) as an MCP sidecar and injects usage instructions into the system prompt
Pi extension: the agent requests secrets via a masked local prompt. Values become env vars for bash, never enter the model context, and are redacted from tool output.
Ports Claude Code's compaction prompt (9-section <analysis>+<summary>) to pi's session_before_compact hook. Faithful reconstruction of the leaked full-compact prompt.
Saves LLM provider request payloads to the project's .pi folder for per-project debugging
Pi extension that transfers context to a new focused session via /handoff, extracting what matters and generating a continuation prompt.
Local coordination layer for AI agent teams: shared backlog, roles, reservations, review state, journal, and prompt context
Make Pi resources optional and context-activated
Transfer context to a new focused pi session via a generated, editable handoff prompt instead of lossy compaction
Run one prompt repeatedly in fresh Pi sessions until its stop condition is met.
Deferred tool loading for Pi: keep heavy tool schemas out of context until prompts need them.
Pi extension to visualize context usage and inspect the hidden parts: base prompt, tools defs, and extension injections
Pi extension for attachment processing — classifies, parses, and renders file attachments into LLM-visible prompt context
Pi extension providing persistent curated memory (MEMORY.md + USER.md) injected into the system prompt as a frozen snapshot.
Pi provider bridge that runs Claude Code through the Claude Agent SDK, with opt-in forwarding for Pi prompt context.
Pi coding-agent extension that summarizes completed tool-call batches, replaces raw outputs with short stubs, compresses closed tool-call chains, and recovers any original on demand via context_tree_query.
Semantic token compression for pi terminal — filters noisy bash output and retroactively masks stale tool results in conversation history (per JetBrains Dec 2025 + Anthropic tool-result clearing)
Pi quality-of-life extension: compact statusline/π prompt, reliable multiline input, styled pasted-image chips, session naming/search/context import, scheduled prompts, handoff, permission prompts, notifications, custom compaction, and a collapsed-thinkin
Codebase-aware prompt enhancer for the Pi coding agent — rewrites rough prompts into precise ones, with project tree, git context, and referenced file contents, before they reach the model.
Deterministic, reversible in-session context compaction for the Pi coding agent — folds stale tool output out of the model's view, never out of the session.
A cozy knowledge home for pi. Distributed as a global package; brain clones are content-only and inherit skills, prompts, themes, tools, and the extension from the install.
Explicit lightweight persistent memory for Pi with no automatic context injection.
Display the full pi coding agent system prompt with injected tools, guidelines, context files, and skills in a scrollable overlay
Compact Pi extension for recording turnlog provenance with low prompt overhead.
Recall your last prompt back to the editor after aborting (Esc)
Custom compaction for pi — swap the model, template, and trigger for context compaction
Model Context Protocol (MCP) integration for pi-coding-agent. Connect to any MCP server (stdio, SSE, StreamableHTTP, WebSocket) and use its tools, resources, and prompts natively.
Memory for pi — Claude-Code-compatible cwd-keyed memory, auto-injected into every prompt. One-command CC→Pi import. The cross-tool on-ramp from Claude Code to pi.
Independent Pi package that exposes backend-backed Wizard tools and prompt context.
Trusted Pi extension binding a Pi session to a reviewr pane: prompt context capture, tool-location reporting, and local review-draft staging over a local socket
Pi extension: resolve @filepath (files & directories) in AGENTS.md, inject into system prompt
Pi extension that reruns a prompt from a clean session checkpoint for bounded Ralph loops.
Model Context Protocol (MCP) integration for pi-coding-agent. Connect to any MCP server (stdio, SSE, StreamableHTTP, WebSocket) and use its tools, resources, and prompts natively.
Pi extension that injects project context into first-turn prompts and compaction continuity messages.
Extensión de Pi que inyecta una carpeta local `.pi/memory/` del proyecto (índice `MEMORY.md` capado más archivos de topic bajo demanda) en el system prompt.
Pi extension that manages conversation context: trims old history, strips stale tool calls, and keeps a rolling summary fed back into the system prompt. Each feature toggles independently.
Intelligent model routing and prompt compression for Pi/Kiro
Developer toolkit for Pi — modular skills for authoring extensions, skills, prompt templates, themes, packages, and context files. Includes introspection tools for version, docs, changelog, benchmarking, and loaded tools.
Config-driven context injection for Pi agents. Loads files at session start and injects them as system prompt or message.
#@file — inject the whole file into your Pi prompt, every time and everywhere, only when you want it. Renders as a compact read-tool line.
Local, read-only prompt provenance inspector for Pi Coding Agent