Continuous Agent Loop

affaan-m/ECC/skills/continuous-agent-loop

by affaan-mef648e01899ba3e8dc6371642deaaf64b4477775No license275K starsListed Oct 9, 2026Updated Oct 9, 2026Repository updated 4 days ago

Patterns for continuous autonomous agent loops with quality gates, evals, and recovery controls. Use when running an agent loop that must self-check, gate on evals, and recover from failures.

Instructions onlyAI & Agents
AI-generated overview

Patterns for running continuous autonomous agent loops with quality gates, evals, and failure recovery.

What it does
This skill provides guidance for selecting and running continuous autonomous agent loops. It offers a decision flow for choosing loop styles such as strict CI/PR control, RFC decomposition, exploratory parallel generation, or sequential execution, and recommends a combined production stack of RFC decomposition, quality gates, an eval loop, and session persistence. It also lists common failure modes and a recovery procedure.
When to use it
Use it when running an agent loop that must self-check, gate on evaluations, and recover from failures. It is suited to teams choosing between loop patterns or diagnosing loop problems such as churn, repeated retries, merge queue stalls, or cost drift.
Requirements
No scripts or special tooling are required; it is an instructions-only skill. It references related skills and commands such as continuous-pr, rfc-dag, infinite, sequential, ralphinho-rfc-pipeline, plankton-code-quality, /quality-gate, eval-harness, nanoclaw-repl, and /harness-audit.

Continuous Agent Loop

This is the v1.8+ canonical loop skill name. It supersedes autonomous-loops while keeping compatibility for one release.

Loop Selection Flow

text
Start  |  +-- Need strict CI/PR control? -- yes --> continuous-pr  |  +-- Need RFC decomposition? -- yes --> rfc-dag  |  +-- Need exploratory parallel generation? -- yes --> infinite  |  +-- default --> sequential

Combined Pattern

Recommended production stack:

  1. RFC decomposition (ralphinho-rfc-pipeline)
  2. quality gates (plankton-code-quality + /quality-gate)
  3. eval loop (eval-harness)
  4. session persistence (nanoclaw-repl)

Failure Modes

  • loop churn without measurable progress
  • repeated retries with same root cause
  • merge queue stalls
  • cost drift from unbounded escalation

Recovery

  • freeze loop
  • run /harness-audit
  • reduce scope to failing unit
  • replay with explicit acceptance criteria

Source and attribution

Source:affaan-m/ECCinskills/continuous-agent-loopat commitef648e0

License: No license

Content belongs to its original authors. SourceWeft indexes it from a public repository.

Report or request removal