Continuous Agent Loop
- 5.7k installs
- 238k repo stars
- Updated August 5, 2026
- affaan-m/everything-claude-code
continuous-agent-loop is an agent skill for sustained autonomous loops with quality gates, evaluation harnesses, and recovery controls when agent sessions churn or stall.
About
The continuous-agent-loop skill is the v1.8+ standard pattern for sustained autonomous agent loops with quality gates, evaluation, and recovery controls, replacing autonomous-loops for one release while preserving compatibility. A loop selection flow routes strict CI and pull request work to continuous-pr, RFC decomposition to rfc-dag, exploratory parallel generation to infinite, and everything else to sequential as the default. Recommended production stacks chain four layers: RFC decomposition via ralphinho-rfc-pipeline, quality gates with plankton-code-quality plus /quality-gate, evaluation loops through eval-harness, and session persistence with nanoclaw-repl. Documented failure modes include loop churn without measurable progress, repeated retries on the same root cause, stalled merge queues, and cost drift from unbounded escalation. Recovery playbooks freeze the loop, run /harness-audit, shrink scope to the failing unit, and replay with explicit acceptance criteria before resuming autonomous work.
- v1.8+ standard loop skill replacing autonomous-loops with one-release compatibility
- Decision tree routes continuous-pr, rfc-dag, infinite, or sequential defaults
- Production stack chains RFC pipeline, quality gates, eval-harness, and nanoclaw-repl
- Failure modes cover churn, repeated root-cause retries, merge stalls, and cost drift
- Recovery freezes loops, runs /harness-audit, shrinks scope, and replays acceptance criteria
Continuous Agent Loop by the numbers
- 5,655 all-time installs (skills.sh)
- +223 installs in the week ending Aug 5, 2026 (Skillselion tracking)
- Ranked #138 of 16,546 AI & Agent Building skills by installs in the Skillselion catalog
- Security screen: LOW risk (skills.sh audit)
- Data as of Aug 5, 2026 (Skillselion catalog sync)
continuous-agent-loop capabilities & compatibility
- Capabilities
- loop selection among continuous pr, rfc dag, inf · production stack pairing rfc pipeline, quality g · failure detection for churn, repeated retries, m · recovery workflow freezing loops, running harnes
- Works with
- github
- Use cases
- orchestration · ci cd · planning
What continuous-agent-loop says it does
これはv1.8+の標準ループスキル名です。1リリースの間、`autonomous-loops`との互換性を保ちながら置き換えます。
+-- Need strict CI/PR control? -- yes --> continuous-pr
測定可能な進捗なしのループチャーン
- `/harness-audit`を実行する
npx skills add https://github.com/affaan-m/everything-claude-code --skill continuous-agent-loopAdd your badge
Show developers this skill is listed on Skillselion. Paste this into your README.
| Installs | 5.7k |
|---|---|
| repo stars | ★ 238k |
| Security audit | 3 / 3 scanners passed |
| Last updated | August 5, 2026 |
| Repository | affaan-m/everything-claude-code ↗ |
How do you choose and operate a production-grade autonomous agent loop without churn, repeated root-cause retries, or runaway cost?
Select and run sustained autonomous agent loops with quality gates, evaluation harnesses, and documented recovery when loops churn or stall.
Who is it for?
Teams running long autonomous coding sessions that need CI-aware loops, RFC decomposition, parallel exploration, or sequential defaults with explicit recovery.
Skip if: Skip for one-shot chat answers, manual edits without harness tooling, or tasks that need only a single tool call with no loop governance.
When should I use this skill?
User sets up sustained agent loops, asks about autonomous-loops migration, loop selection, production harness stacks, or recovery from agent churn.
What you get
A routed loop mode plus a stacked harness with quality gates, evaluation, session persistence, and a documented freeze-audit-shrink-replay recovery path.
- Selected agent loop topology configuration
- Quality gate and recovery control definitions
By the numbers
- v1.8+ canonical skill superseding autonomous-loops for one release cycle
- Documents 4 loop topology branches: continuous-pr, rfc-dag, infinite, and sequential
Files
継続的エージェントループ
これはv1.8+の標準ループスキル名です。1リリースの間、autonomous-loopsとの互換性を保ちながら置き換えます。
ループ選択フロー
Start
|
+-- Need strict CI/PR control? -- yes --> continuous-pr
|
+-- Need RFC decomposition? -- yes --> rfc-dag
|
+-- Need exploratory parallel generation? -- yes --> infinite
|
+-- default --> sequential組み合わせパターン
推奨される本番スタック： 1. RFC分解（ralphinho-rfc-pipeline） 2. 品質ゲート（plankton-code-quality + /quality-gate） 3. 評価ループ（eval-harness） 4. セッション永続化（nanoclaw-repl）
失敗モード
- 測定可能な進捗なしのループチャーン
- 同じ根本原因での繰り返しリトライ
- マージキューの停止
- 無制限のエスカレーションによるコストドリフト
リカバリー
- ループを凍結する
/harness-auditを実行する- スコープを失敗ユニットに縮小する
- 明示的な受け入れ基準でリプレイする
Related skills
Setup Matt Pocock SkillsScaffold the per-repo configuration that Matt Pocock’s engineering agent skills rely on so they understand the issue tracker, triage labels, and domain documentation la495k196k
Lark AppsConnect Claude, Cursor or custom agents directly to Lark (Feishu) for messaging, document automation, approval workflows and enterprise data access.402k
CavemanSlash token usage by roughly 75% while keeping every technical detail intact when working with Claude Code, Cursor or similar agents.391k94.7k
Lark Skill MakerQuickly turn any Lark/Feishu OpenAPI call or multi-step workflow into a reusable agent skill with its own SKILL.md.386k16k
Running Claude Code Via Litellm CopilotRun Claude Code at a fraction of the cost by routing requests through LiteLLM to the GitHub Copilot Chat API.270k72
Codex PetGenerate a complete Codex Pet spritesheet and metadata from one reference image without needing an OpenAI key or Codex Pro.246k10Forks & variants (1)
Continuous Agent Loop has 1 known copy in the catalog totaling 1.4k installs. They canonicalize to this original listing.
- affaan-m - 1.4k installs
How it compares
Pick continuous-agent-loop over one-shot agent prompts when runs need explicit loop topology selection, eval gates, and recovery between iterations.
FAQ
Which loop mode should I pick first?
Follow the selection flow: continuous-pr for strict CI or PR control, rfc-dag for RFC decomposition, infinite for exploratory parallel generation, otherwise sequential.
What production stack does the skill recommend?
Chain ralphinho-rfc-pipeline, plankton-code-quality with /quality-gate, eval-harness, and nanoclaw-repl for session persistence.
How do I recover when a loop keeps failing?
Freeze the loop, run /harness-audit, shrink scope to the failing unit, and replay with explicit acceptance criteria.
Is Continuous Agent Loop safe to install?
skills.sh reports 3 of 3 security scanners passed. Review the Security Audits panel on this page before installing in production.