Free personal project · Proof of product engineering capability

An AI assistant that truly gets work done.

Slip Loomer covers both office documents and software development: Ask is read-only understanding; Agent edits files, runs commands, and calls tools.

Contact：Study ExchangeEmail：StudyExchange@163.com

VS Code 1.85+ / Cursor · Bring your own model API

SLIP LOOMERWorkspace

Analyze sales.xlsx, identify anomalies, and create a presentation PPT

Agent is working

read_office_file sales.xlsx

Analyze regional and SKU anomalies

write_office_file report.pptx

Tell Slip Loomer what to do next…

Free personal project · Proof of capability

Slip Loomer

A VS Code / Cursor AI assistant that truly reads documents, edits code, and runs tools.

From Agent orchestration and security boundaries to Office files and multi-session use, Slip Loomer is a complete AI Agent product-engineering practice, and it shows the system skills needed for enterprise-field Agents.

Read-only Ask Q&A / automated Agent execution
Word, Excel, PPT, and PDF processing
Code edits, Diff review, and LSP diagnostics
Skills, Memory, MCP, and web tools
Bring your own OpenAI-compatible model API
PathGuard workspace security boundaries

View on GitHub Eligible individuals may use it for free under the current license. Welcome to learn, exchange ideas, and grow together.

Core capabilities

More than conversation—it gets work done.

From model orchestration and file security to Office documents and software engineering, all capabilities come together in one workspace.

Dual Ask and Agent modes

When you only need understanding, no files are changed. When execution is needed, Agent plans the work, invokes tools, and presents the changes.

Office document processing

Read, create, and edit Word, Excel, and PowerPoint files, with PDF conversion to PPT, Word, or Markdown.

Software development workflows

Search code, edit files, review Diffs, run terminal commands, and read LSP diagnostics across development and troubleshooting.

Clear security boundaries

PathGuard limits readable and writable paths; gitignore paths cannot be read or written, and sensitive files never enter the context.

Skills and Memory

Load project Skills on demand, retain local memory across sessions, and automatically compress context in long conversations.

MCP and bring-your-own models

Connect custom MCP tools, use OpenAI-compatible APIs, and assign different models to different roles.

Example use cases

Documents and code both become real files.

These are representative workflows already supported by Slip Loomer.

Office documents

Create a presentation PPT from a sales Excel file

Read sales.xlsx, analyze regional, product, and time trends, then generate report.pptx with conclusions.

Output: analysis conclusions + an editable PPT file

Software development

Locate and fix a post-login 500 error

Search auth-related code, inspect the call chain, edit the affected files, present the Diff, and run tests to verify the fix.

Output: code changes + diagnostic and test results

Hybrid tasks

Update the design and implementation from web sources

Find and retrieve official documentation while updating the repository README, technical-design Word document, and sample code.

Output: source summary + documents + code

Project status

156 test files and 732 test cases.

Covering Agent orchestration, Office/PDF, code tools, PathGuard, Memory, MCP, and representative user scenarios.

View source and documentation