issue-driven-al
Orchestrates issue-driven agent loops that post cycle status to issue threads and incorporate human feedback in each cycle.
What this skill does
<!-- AIWG-SKILL-CALLOUT --> > **Skill access pattern (post-kernel-pivot, 2026.5+)** > > Skill names referenced in this document are AIWG skills, **not slash commands**. Most are not kernel-listed and cannot be invoked as `/skill-name` by the platform. Reach them via: > > ```bash > aiwg discover "<capability>" > aiwg show skill <name> > ``` > > Only kernel-listed skills (`aiwg-doctor`, `aiwg-refresh`, `aiwg-status`, `aiwg-help`, `use`, `steward`) are directly invokable as slash commands. See [skill-discovery rule](../../../addons/aiwg-utils/rules/skill-discovery.md). # Issue-Driven Agent Loop Skill ## Triggers Alternate expressions and non-obvious activations (primary phrases are matched automatically from the skill description): - "work through the bugs" → issue-driven loop shorthand - "address the backlog" → iterative issue queue processing - "tackle issue [N]" → single-issue loop ## Purpose Transforms the issue tracker from a passive record into an active 2-way collaboration surface. Each Al cycle posts structured status to the issue thread, scans for human feedback, and responds substantively. The human can monitor and steer agent work asynchronously by commenting on the issue. ## Behavior When triggered, this skill invokes the `address-issues` command with appropriate parameters extracted from the user's natural language request. ### Parameter Extraction | User Says | Extracted Parameters | |-----------|---------------------| | "address issues 17 18 19" | `17 18 19` | | "fix open bugs" | `--filter "status:open label:bug"` | | "work on the bug backlog" | `--filter "status:open label:bug"` | | "tackle issue 17" | `17` | | "go through all open issues" | `--all-open` | | "address the open issues interactively" | `--all-open --interactive` | | "focus on security bugs" | `--all-open --guidance "Focus on security bugs"` | | "fix bugs 17-19, they're all auth related" | `17 18 19 --guidance "These are all related to auth"` | ### Parameter Support This skill supports the standard AIWG `--interactive` and `--guidance` parameters: **`--interactive`**: When detected in user intent (e.g., "interactively", "walk me through"), adds `--interactive` to invoke discovery questions before starting and pause between issues for human go/no-go. **`--guidance`**: When the user provides upfront direction (e.g., "focus on bugs", "skip feature requests", "security first"), extracts the guidance text and passes it via `--guidance "..."` to tailor prioritization and approach without interactive prompts. ### Execution Steps 1. **Parse intent** — identify issue numbers, filters, or "all open" from user message 2. **Invoke `address-issues`** with extracted parameters 3. **Monitor execution** — the command handles the full cycle protocol ### The 3-Step Cycle Protocol Each cycle of the agent loop follows this protocol: **Step 1: Work** — Read issue context, implement fix/feature, run tests **Step 2: Post Status** — Post structured comment to the issue thread: ``` **AL CYCLE #N – [Progress|Blocked|Review Needed]** - Actions taken this cycle - Task checklist (updated) - Blockers (if any) - Next steps ``` **Step 3: Scan & Respond** — Read all new thread comments, classify them (feedback/question/approval/correction), and incorporate into the next cycle. Never ignore human input. ### Thread Scanning Classification | Comment Type | Agent Response | |-------------|---------------| | Feedback | Incorporate into next cycle's work | | Question | Answer in next status comment | | Approval | Proceed to next phase or close issue | | Correction | Adjust approach, acknowledge the change | | Automated/bot | Ignore | ### Completion Criteria (per issue) An issue is resolved when: - Implementation is complete - Tests pass - Documentation updated (if needed) - All thread feedback addressed - No unresolved blocker comments ### Multi-Issue Strategy | Strategy | Trigger | |----------|---------| | Sequential | Default — one issue at a time | | Batched | When user mentions "related" issues or same module | | Parallel | When user says "in parallel" (respects context budget) | ## Integration ### Issue Tracker APIs **Gitea** (via MCP tools): - `mcp__gitea__list_repo_issues` — list issues - `mcp__gitea__get_issue_by_index` — read issue - `mcp__gitea__get_issue_comments_by_index` — read thread - `mcp__gitea__create_issue_comment` — post status - `mcp__gitea__edit_issue` — update labels/status **GitHub** (via `gh` CLI): - `gh issue list` — list issues - `gh issue view N` — read issue - `gh issue comment N --body "..."` — post status - `gh issue close N` — close resolved ### Related Commands | Command | Integration | |---------|-------------| | `ralph` | Core loop engine | | `issue-list` | Fetches issues by filter | | `issue-comment` | Posts cycle status comments | | `issue-close` | Closes resolved issues | | `issue-sync` | Links commits to issues | ## Safety - Never force-push or make destructive git changes - Always run tests before posting completion status - Respect `--max-cycles` limit (default: 6) - Post status every cycle — transparency is mandatory - On error, post blocker comment rather than failing silently - In `--interactive` mode, pause between issues for human go/no-go ## Configuration Default settings (overridable via parameters): - Max cycles per issue: 6 - Strategy: sequential - Provider: auto-detect from project config - Branch per issue: false - Interactive: false ## Examples ### Fix a specific bug ``` User: "tackle issue 17" → /address-issues 17 ``` ### Work through the bug backlog ``` User: "work on the bug backlog" → /address-issues --filter "status:open label:bug" ``` ### Interactive session on multiple issues ``` User: "address issues 17, 18, 19 interactively" → /address-issues 17 18 19 --interactive ``` ### All open issues with higher cycle limit ``` User: "go through all open issues, give each one up to 8 cycles" → /address-issues --all-open --max-cycles 8 ``` ### With guidance ``` User: "fix the open bugs, focus on security issues first" → /address-issues --filter "status:open label:bug" --guidance "Security issues are top priority" ``` ### Guidance with batch context ``` User: "address issues 17, 18, 19 — they're all related to the auth refactor" → /address-issues 17 18 19 --guidance "These are all related to the auth refactor, address them as a batch" ``` ## References - @.aiwg/planning/issue-driven-ralph-loop-design.md - Design document - @$AIWG_ROOT/agentic/code/frameworks/sdlc-complete/skills/address-issues/SKILL.md - Command definition - @$AIWG_ROOT/agentic/code/addons/ralph/skills/ralph/SKILL.md - Agent loop command - @$AIWG_ROOT/agentic/code/frameworks/sdlc-complete/skills/issue-auto-sync/SKILL.md - Issue sync skill - @$AIWG_ROOT/agentic/code/addons/aiwg-utils/rules/context-budget.md - Parallel subagent limits
Related in AI Agents
skill-development
IncludedComprehensive meta-skill for creating, managing, validating, auditing, and distributing Claude Code skills and slash commands (unified in v2.1.3+). Provides skill templates, creation workflows, validation patterns, audit checklists, naming conventions, YAML frontmatter guidance, progressive disclosure examples, and best practices lookup. Use when creating new skills, validating existing skills, auditing skill quality, understanding skill architecture, needing skill templates, learning about YAML frontmatter requirements, progressive disclosure patterns, tool restrictions (allowed-tools), skill composition, skill naming conventions, troubleshooting skill activation issues, creating custom slash commands, configuring command frontmatter, using command arguments ($ARGUMENTS, $1, $2), bash execution in commands, file references in commands, command namespacing, plugin commands, MCP slash commands, Skill tool configuration, or deciding between skills vs slash commands. Delegates to docs-management skill for official documentation.
reprompter
IncludedTransform messy prompts into well-structured, effective prompts — single or multi-agent. Use when: "reprompt", "reprompt this", "clean up this prompt", "structure my prompt", rough text needing XML tags and best practices, "reprompter teams", "repromptception", "run with quality", "smart run", "smart agents", multi-agent tasks, audits, parallel work, anything going to agent teams. Don't use when: simple Q&A, pure chat, immediate execution-only tasks. See "Don't Use When" section for details. Outputs: Structured XML/Markdown prompt, quality score (before/after), optional team brief + per-agent sub-prompts, agent team output files. Success criteria: Single mode quality score ≥ 7/10; Repromptception per-agent prompt quality score 8+/10; all required sections present, actionable and specific.
adaptive-compaction
IncludedAdaptive add-on policy and recovery layer that decides WHEN to compact, prune, snapshot, or fork -- replacing fixed-percent auto-compaction across Claude Code, Codex, and MCP-capable hosts. Trigger on auto-compact timing or damage: "when should I compact", "is it safe to compact now or start a fresh session", "auto-compact fires too early/mid-task", "switching to an unrelated task but the window still has space", "context rot", "answers get worse the longer the session runs", "the agent forgot the plan or my decisions after it summarized", "add a layer on top that manages context without changing the agent", raising autoCompactWindow to give the policy room, or installing/tuning a cross-tool compaction policy or PreCompact hook -- even when "compaction" is never said but the problem is context-window pressure or post-summarization memory loss. Do NOT use to summarize a conversation, build RAG, write a summarization prompt (decides WHEN not HOW), or answer max-context-length trivia.
agent-skill-creator
IncludedCreate cross-platform agent skills from workflow descriptions. Activates when users ask to create an agent, automate a repetitive workflow, create a custom skill, or need advanced agent creation. Triggers on phrases like create agent for, automate workflow, create skill for, every day I have to, daily I need to, turn process into agent, need to automate, create a cross-platform skill, validate this skill, export this skill, migrate this skill. Supports single skills, multi-agent suites, transcript processing, template-based creation, interactive configuration, cross-platform export, and spec validation.
llm-wiki
IncludedUse when building or maintaining a persistent personal knowledge base (second brain) in Obsidian where an LLM incrementally ingests sources, updates entity/concept pages, maintains cross-references, and keeps a synthesis current. Triggers include "second brain", "Obsidian wiki", "personal knowledge management", "ingest this paper/article/book", "build a research wiki", "compound knowledge", "Memex", or whenever the user wants knowledge to accumulate across sessions instead of being re-derived by RAG on every query.
skill-master
IncludedAgent Skills authoring, evaluation, and optimization. Create, edit, validate, benchmark, and improve skills following the agentskills.io specification. Use when designing SKILL.md files, structuring skill folders (references, scripts, assets), ingesting external documentation into skills, running trigger evals, benchmarking skill quality, optimizing descriptions, or performing blind A/B comparisons. Keywords: agentskills.io, SKILL.md, skill authoring, eval, benchmark, trigger optimization.