ECS-Discord-Bot

Safe, project-wide remediation of a dangerous or incorrect code pattern using structured search and replace.

Run a multi-faceted code review on uncommitted changes using specialized review subagents.

Create an Architectural Decision Record (ADR) to document a design choice or technical strategy.

Run a multi-faceted code review on uncommitted changes using specialized review subagents.

Commit (if needed), push the current branch to origin, and generate a pull request description in the chat.

Generate a session handoff document capturing the current working state for the next session.

Proactively audit GitHub Actions workflows and CI/CD logs for deprecations, stale runtimes, and missing best practices.

Orchestrate an automated implement → review → fix cycle for tasks in a spec. Chains the `implement-task` and `review-code` skills in a loop until code is clean.

Automate the Bedrock AgentCore Runtime lifecycle (deploy, restore environment variables, verify health).

Analyze local telemetry logs to produce metrics on skill usage, agent invocations, and workflow patterns.

Capture learnings, patterns, and corrections discovered during this session into persistent project memory. This runs automatically at the end of `session-handoff` and can be invoked standalone.

Build a container image for linux/amd64 (Fargate), push to ECR, and rotate the ECS Fargate service.

Create a new Kiro CLI skill (prompt) from a conversation, a pasted prompt, or a description.

Wrap up a completed issue: verify all tasks done, group deferred findings, deduplicate against existing GH issues, create follow-up issues, close with summary, move spec to Done.

Orchestrate the full specification pipeline — requirements → high-level design → low-level design → task plan — producing a single unified spec document. This spec is the primary input for `implement-and-review-loop`.

Expert front-end designer and brand strategist for the ExtraLife.Web.Admin Angular application. Use this skill when making UI/UX changes, defining visual patterns, or migrating the interface to a new design system.

Produce a high-level architecture design from a requirements document.

Produce a detailed component-level design from a high-level design document.

Post-deploy visual review of the live application. Catalog UX issues, run contrast checks, and log findings to the spec.

Given a GitHub Actions run URL or run ID, fetch the failed job logs, identify root cause, and suggest a fix. Use this whenever a CI/CD workflow fails.

Systematically review AI chat traces to find failure patterns using error analysis methodology (inspired by Hamel Husain's evals framework).

Deterministic pre-commit and post-implementation checks that simulate hooks. Run automatically at the end of `implement-task` Phase 5 (verify build) and before any commit.

Orchestrate an automated implement → review → fix cycle for tasks in a spec. Chains the `implement-task` and `review-code` skills in a loop until code is clean.

Implement a specific task from the task plan using Test-Driven Development methodology, then verify documentation is current.

Review the current chat session where a skill was used and improve it based on feedback.

Systematic investigation of a technical blocker using tools to verify assumptions, document evidence, and create escalation artifacts for external teams.

Run a side-by-side model evaluation for an agent, comparing two Bedrock models on identical queries.

Create an implementation task plan with status tracking and dependency graph from a design document.

Fast checklist-based code review of uncommitted changes. Use this for a quick pass; use `review-code` for the full multi-agent review.

Capture a stream-of-consciousness description of requirements and produce a structured requirements document.

Evaluate an AWS service or feature for adoption. Produces a structured recommendation with region availability, architecture mapping, and a file-or-skip decision.

Detect user intent and automatically chain to the correct skill. This reduces the "which skill do I run?" friction and moves toward auto-activated skills.

Generate and present SQL validation queries for a new or modified stored procedure. The user runs the queries against RDS and reports results back.

Run three secret-scanning tools against the codebase to detect credentials, API keys, tokens, and other sensitive data before committing.

Resume a previous working session by loading the latest handoff context.

Verify that code changes have been successfully deployed to the target environment after CI/CD completes.

Validate the contract between skill phases to catch broken handoffs early. This is the "schema validation at agent boundaries" pattern from distributed systems.