Try-Works

Try-Works

An index of projects, products, and ideas.<br>Projects & Products

rlm-workflow<br>tinytunes DJ

lurkkit<br>role-model<br>Pocketmodel

Writing

recursive-mode for coding agents<br>Why Chinese AI labs went open and will remain open

Coding agents and the growing 1% problem

role-model: the case for a model routing protocol

Contact

Email

My new Twitter

role-model

Model routing protocol<br>Router<br>runtime

Pi package

Protocol and runtime for hybrid local/API model inference. Models are assigned roles based on capabilities, performance, speed and cost; task requests are routed to the best suited models to optimiza the triangle of constraints, cost-speed-accuracy.

Github: https://github.com/try-works/role-model<br>Docs: https://role-model.dev/

rlm-workflow

Agent Skill for Codex, etc

Improved context length, code quality, and traceability

Installation<br>npx skills add https://github.com/doubleuuser/rlm-workflow --skill rlm-workflow

skills.sh

https://skills.sh/doubleuuser/rlm-workflow/rlm-workflow

GitHub<br>https://github.com/doubleuuser/rlm-workflow

After the MIT paper Recursive Language Models

paper demonstrated a method of increasing effective context length to 10M tokens by using sub-agents to move information from the context window to an information store outside the chat, there's been a number of different takes on how to put this into practice in development workflows. Some even go in the direction of storing entire session contexts in a database for later retrieval to preserve reasoning for changes that were made.

rlm-workflow is yet another take, this time with a slightly different angle:

Important information like requirements, codebase analysis and implementation plans should not be passed in the chat in the first place. Chat is effectively CLI and should be used for invocations and commands, not for passing information.

rlm-workflow is modelled after a regular kanban workflow from requirement to implementation plan to testing and manual QA. The workflow is sequential and phase; each phase outputs one markdown doc and takes the previous phases' docs as input. Each phase is gated on fulfilling criteria defined in the previous phase, and at the end of a phase, its output docs are locked.

The user first creates the 00-requirements.md doc in an RLM folder, then invokes the workflow in chat. It then runs until the manual QA stage, where it waits for user approval before continuing. After finishing an RLM run, the agent updates DECISIONS.md which is a ledger of requirements implemented previously, their whys and whats, and links to respective RLM docs. It also updates STATE.md, an overview of the app's current state.

To be practical, this is what your repo will look like:

rlm/00-my-first-requirements/<br>- 00-requirements.md (user-created)<br>- 01-as-is.md<br>- 02-to-be.md<br>- 03-implementation-summary.md<br>- 04-manual-qa.md (test cases are pre-defined; the user enters pass/fail and notes in the doc)<br>+ /addenda/ if needed

To summarize:<br>1. Specs are never passed through the chat so they do not suffer from context rot<br>2. Work is always done based on docs that are locked, so it cannot suffer from degradation<br>3. The workflow is self-documenting; it is also easily human readable; can also be used to generate information for non-technical stakeholders<br>4. There is no need to index the codebase a database. The rlm docs provide progressive disclosure and point the model in the right direction. Should significantly reduce token usage.<br>4. In my simple test, the workflow improves both quality and time to success for complex requirements

The benefits of using rlm-workflow for assisted engineering includes improved traceability through workflow and global docs, reduced token usage, reduced context rot, improved accuracy and code quality, and improved speed, when considering reduced need for re-work.

What about the chat sessions? Forget about them. Instructions don't matter, only outcomes.

tinytunes DJ

minimal DJ controller and music player

Software, web-based

tinytunes DJ is a lightweight DJ controller that runs in your browser, on your laptop or desktop, tablet or smartphone device, so that you’re ready to play anywhere you go.

It is meant as a casual, zero-friction music service that is available anywhere, all the time, for DJs to discover new music, compose playlists and set lists, and enjoy music on the go by using the autoplay toggle to turn the controller into a music player.

https://dj.t-tunes.com/

lurkkit

Meta UI for reddit, substack

Web application

lurkkit is a meta UI that allows users to build their own frontpage feeds based on subreddits and substack publications. reddit and substack support viewing the full post inline in the feed, while reddit sources also have viewable comments.

reddit and substack sources can be sorted into categories, ranking can be boosted per category and individual source, and a specific priority feed can be created for sources and publications that shouldn’t be...