Compile skill capabilities and goals into executable JSON sequence files.
npx skills add https://github.com/jnesfield-bot/rho --skill skill-sequencer
Please help me install this Agent Skill. Skill: skill-sequencer Source: https://github.com/jnesfield-bot/rho/tree/main/skills/skill-sequencer Command: npx skills add https://github.com/jnesfield-bot/rho --skill skill-sequencer
This Skill transforms a flexible, LLM-driven skill into a precise, deterministic sequence of steps, making complex tasks auditable, editable, and reliably repeatable.
arxiv-research skill to find and implement a specific algorithm from a given paper ID, creating a reusable recipe for future similar tasks.Compile the arxiv-research skill for the goal "Find and implement the DQN algorithm from arXiv:1312.5602" and save it to ./sequences/implement-dqn.json.
High-intent search queries and answers about installing and using this skill.
You compile an LLM-driven skill into a deterministic JSON sequence by defining the skill's capabilities and a specific goal, which translates flexible agent actions into a precise, executable step-by-step format.
A deterministic sequence file is an executable JSON recipe that translates LLM-driven skill descriptions into step-by-step procedural formats, enabling auditable and reliably repeatable task execution without re-planning.
Agents re-plan steps because they lack a deterministic sequence file; compiling the skill capabilities into a procedural JSON recipe enables execution of pre-defined steps without requiring the LLM to re-plan each time.
Yes, you can review, modify, and version control the exact steps an automated agent takes because the deterministic JSON sequence file is auditable and editable, supporting variable substitution and on-failure policies.
No external dependencies are required to generate a repeatable task recipe; the skill uses internal scripts to compile capabilities and goals into a standalone, executable JSON sequence file.
Variable substitution and on-failure policies are built into the compiled JSON sequence file, allowing you to define robust workflow automation parameters directly when translating skill descriptions into steps.
liqiongyuConvert unstructured skill descriptions into standardized agent-executable skill packs.
hubvueChain multiple skills sequentially, passing outputs as inputs.
refly-aiManage GitHub repositories, issues, and pull requests programmatically.
illogicalPlan LLM project architectures with a five-stage pipeline and file-system state tracking.