sijeeshmiziha

sijeeshmiziha/visionagent

Command Claude Code

Scaffold and implement a new tool for VisionAgent.

sijeeshmiziha/visionagent

Command Claude Code

Diagnose and fix a bug or failing test in VisionAgent.

sijeeshmiziha/visionagent

Command Claude Code

You are in full autonomous implementation mode. Follow the VisionAgent agentic task approach from CLAUDE.md rigorously.

sijeeshmiziha/visionagent

Command Claude Code

Create a well-structured PR for the current branch's changes.

sijeeshmiziha/visionagent

Command Claude Code

Prepare and publish a new release of VisionAgent to npm.

sijeeshmiziha/visionagent

Command Claude Code

Perform a thorough code review of changes in this branch.

sijeeshmiziha/visionagent

Command Claude Code

Run tests intelligently based on what changed, then fix any failures.

sijeeshmiziha/visionagent

Hook Claude Code

Runs when the agent finishes a response, running an inline shell check. From sijeeshmiziha/visionagent.

sijeeshmiziha/visionagent

Hook Claude Code

Runs after a tool call finishes for Bash tool calls, running true. From sijeeshmiziha/visionagent.

sijeeshmiziha/visionagent

Settings file Claude Code

Agent settings declaring 2 hook events (Stop, PostToolUse).

sijeeshmiziha/visionagent

Cursor rule Cursor

Agentic workflow patterns—when to use agents vs workflows, prompt chaining, routing, evaluator-optimizer.

sijeeshmiziha/visionagent

Cursor rule Cursor

Coding agents—prompt pattern for code tasks, error-proofing file/edit tools, minimal scaffolding.

sijeeshmiziha/visionagent

Cursor rule Cursor

Context engineering for agents—finite context, prompt altitude, compaction, note-taking, sub-agents.

sijeeshmiziha/visionagent

Cursor rule Cursor

Design patterns (GoF) and file size conventions for sweagent TypeScript codebase.

sijeeshmiziha/visionagent

Cursor rule Cursor

Best practices for long-running agents (initializer/coding pattern, incremental progress, clean state).

sijeeshmiziha/visionagent

Cursor rule Cursor

Multi-agent systems—orchestrator-worker, delegation, scaling, evaluation, production.

sijeeshmiziha/visionagent

Cursor rule Cursor

When and how to use a "think" tool or extended thinking for complex tool-use agents.

sijeeshmiziha/visionagent

Cursor rule Cursor

Tool design for agents—self-contained tools, minimal sets, token-efficient results, progressive disclosure.

sijeeshmiziha/visionagent

Cursor rule Cursor

Evaluation-driven tool improvement—prototype, evaluate, analyze, collaborate with agents.

sijeeshmiziha/visionagent

Instructions file

Instructions for sijeeshmiziha/visionagent, covering visionagent — claude code guide, project overview, architecture, key abstractions and development workflow.