Agent skill
dev-ralph
Persistence loop — keep working on a task until it's resolved or until a circuit breaker stops you. Different from /loop (which is time-based) — ralph is iteration-based with state.
Install this agent skill to your Project
npx add-skill https://github.com/EvolutionAPI/evo-nexus/tree/main/.claude/skills/dev-ralph
SKILL.md
Dev Ralph
Derived from oh-my-claudecode (MIT, Yeachan Heo). Adapted for the EvoNexus Engineering Layer.
Persistence loop. Keep iterating on a task until it's resolved, with a circuit breaker for safety. Different from /loop (which runs a prompt every N minutes) — ralph runs the same agent repeatedly until a goal is met.
Use When
- Task has a clear completion signal but unclear iteration count
- "Fix until tests pass" type work
- Auto-correcting loops where each iteration uses the previous result
Do Not Use When
- Time-based polling → use
/loop(built-in) - Single-shot work
- No clear completion signal (you'll loop forever)
Iron Rules
- Completion signal must be defined upfront — "tests pass" or "build green" or "verifier APPROVES"
- Circuit breaker: stop after 5 iterations OR after 3 identical failures
- State persisted between iterations — what's been tried, what failed, why
- Never silently spin — emit progress every iteration
Workflow
Iteration 1:
- Run agent (e.g., @bolt-executor) with task
- Check completion signal
- If met: STOP, success
- If not: record what was tried, what failed
- Increment iteration counter
Iteration 2:
- Run agent with task + accumulated state
- Check completion signal
- ...
Iteration 5 (max):
- Stop and report: "Reached max iterations without completion"
- Hand off to @apex-architect with full state dump
OR
3-identical-failure breaker:
- If iteration N produces the same error fingerprint as N-1 and N-2
- Stop immediately
- The issue is no longer "fix the bug" — it's "question the approach"
Output
Per iteration, emit:
[ralph iteration {N}/{max}]
Goal: {completion signal}
Tried: {what this iteration did}
Result: {success | failure: {reason}}
State: {what's been tried so far}
Final report saved to workspace/development/debug/[C]ralph-{topic}-{date}.md.
Pairs With
@bolt-executor(most common driver)@hawk-debugger(for bug-fix loops)@oath-verifier(provides the completion signal)@apex-architect(escalation when circuit breaker trips)
Cancellation
Interrupted by dev-cancel — preserves state for resume.
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
data-create-viz
Cria visualizações de dados de qualidade profissional com Python no tema Evolution (fundo escuro, acento
cs-customer-escalation
Empacota uma escalação para Devs, Produto ou Davidson com contexto completo. Use quando um bug precisa de atenção além do suporte normal, vários clientes reportam o mesmo problema, um cliente está ameaçando cancelar, ou um problema ficou sem resolução além do SLA. / Package an escalation for engineering, product, or leadership with full context. Use when a bug needs engineering attention beyond normal support, multiple customers report the same issue, a customer is threatening to churn, or an issue has sat unresolved past its SLA.
mkt-draft-content
Draft blog posts, social media, email newsletters, landing pages, press releases, and case studies with channel-specific formatting and SEO recommendations. Use when writing any marketing content, when you need headline or subject line options, or when adapting a message for a specific platform, audience, and brand voice.
gog-tasks
Create, manage, and prioritize tasks and todo items. Convert emails to tasks, set priorities (P0-P3) and categories (Work/Personal/Errands/Admin), review daily priorities, track blocked and overdue tasks. Use when user mentions tasks, todos, action items, or wants to convert emails to tasks. Requires confirmation before creating or deleting tasks.
discord-get-messages
Retrieve messages from Discord channels via the Discord API. Use this skill when the user wants to read, search, or analyze messages from a Discord channel.
cs-customer-research
Pesquisa multi-fonte sobre pergunta ou tópico de cliente com atribuição de fontes. Use quando um cliente pergunta algo que precisa ser verificado, investigando se um bug foi reportado antes, verificando o que foi dito anteriormente a uma conta específica, ou coletando contexto antes de redigir uma resposta. / Multi-source research on a customer question or topic with source attribution. Use when a customer asks something you need to look up, investigating whether a bug has been reported before, checking what was previously told to a specific account, or gathering background before drafting a response.
Didn't find tool you were looking for?