Inject contextual rules into prompts via an 8-layer pipeline.
npx skills add https://github.com/FelipeMaines/ProjetoWeb --skill synapse-felipemaines
Please help me install this Agent Skill. Skill: synapse Source: https://github.com/FelipeMaines/ProjetoWeb/tree/main/.claude/skills/synapse Command: npx skills add https://github.com/FelipeMaines/ProjetoWeb --skill synapse-felipemaines
SYNAPSE helps you consistently inject the right contextual rules into every prompt so your agent follows architecture, domains, and workflows without manual repetition.
<synapse-rules> XML block appended to each prompt.*synapse sub-commands, including creating custom domains and CRUD rule operations.*synapse status and *synapse debug to inspect loaded domains, bracket state, and rule counts.Ask the AI to run *synapse status to see the current SYNAPSE engine state, active domains, and the current context bracket.
High-intent search queries and answers about installing and using this skill.
You can inject contextual rules automatically by using a context engine that appends an XML block to each prompt. This ensures consistent rule application across agent, workflow, task, and squad scenarios without manual repetition.
Token budgeting adapts injection volume based on context brackets like FRESH, MODERATE, DEPLETED, and CRITICAL. This bracket-aware approach automatically scales the injected rules to avoid overrunning the available context window.
The best way to manage context domains is through dedicated sub-commands that support CRUD operations and custom domain creation. This allows you to structure rule flows across multiple layers, from Constitution to Star-Commands.
You can debug rule injection by running status and debug commands to inspect loaded domains, bracket states, and rule counts. This helps identify whether rules are properly loaded or if token budgeting is limiting injection.
Yes, prompt rules can be scoped to specific tasks, agents, workflows, and squads. The context engine also supports keyword-triggered recall and explicit star-commands to activate rules only when relevant.
The 8-layer rule pipeline outputs a structured XML block appended to each prompt. This format ensures the injected rules are clearly delimited and machine-readable for the agent's context window.
specd-sddTranslate user intent into a SpecD change plan with exploration context.
hyagovianaManage SYNAPSE domain rules and debug the 8-layer context injection pipeline.
moasadiValidate DDD layer separation, dependency rules, and domain patterns in codebases.
bacocoGenerate prioritized Synapse skills from Cortex memory, PRD files, and task lists.