Architecture & codebase
12,711 rows, by stacks then stars
1,011 of these skill files have been read, by 858 distinct authors, and what they tell an agent counted. What Architecture & codebase authors agree on, and what they forbid
vaibhavsaxena022/skills/explore-codebase Skill
0★ repoUnderstand the relevant part of a codebase BEFORE changing it. Use at the start of any bug fix or feature in an unfamiliar area. Maps the entry points, data flow, and conventions, and pinpoints where the change belongs — without modifying anything.
vaibhavsaxena022/skills/idempotency Skill
0★ repoMake a mutating endpoint safe to retry. Use when a create/POST or payment-like operation could arrive twice (client retries, at-least-once queues). Adds idempotency keys so duplicates don't double-apply.
vaibhavsaxena022/skills/junit-test-scaffold Skill
0★ repoGenerate JUnit 5 test skeletons for a Java class or method. Use when asked to add tests, scaffold tests, or improve coverage for Spring/Java code. Produces arrange-act-assert tests with mocked dependencies and edge cases.
vaibhavsaxena022/skills/layered-architecture Skill
0★ repoEnforce clean Controller → Facade → Service → Repository layering in a service. Use when adding or reviewing code to keep responsibilities in the right layer and prevent logic or entities from leaking across boundaries.
Paldom/landing-page-builder-skills/skills/nextjs-landing-page Skill
0★ repoImplements a landing page in Next.js App Router with React + shadcn/ui + Tailwind - RSC/client-component boundaries, shadcn setup, next/image & next/font, and the streaming-metadata gotcha. Use when asked to build, scaffold, or fix a Next.js landing page, resolve 'use client' errors, set up shadcn, or fix a blank-head/unindexed page. Not for copy, visual design, animation, or SEO strategy.
grimoire-rs/grimoire-vscode/.claude/skills/swarm-plan Skill
0★ repoUse for feature planning, task decomposition, multi-perspective research, or ADR scaffolding. Tier (`low | auto | high | max`) scales research depth, architect model, and review breadth. Canonical research primitive for AI config / ADR work.
SamyakJhaveri/loam/cultivation/marketplace/pocock-engineering/skills/improve-codebase-architecture Skill
0★ repoFind deepening opportunities in a codebase, informed by the domain language in DOMAIN.md and the decisions in docs/adr/. Use when the user wants to improve architecture, find refactoring opportunities, consolidate tightly-coupled modules, or make a codebase more testable and AI-navigable.
SamyakJhaveri/loam/cultivation/marketplace/understand-anything/skills/understand-anything Skill
0★ repoMulti-agent codebase analysis with interactive React knowledge graph dashboard. Best for onboarding and visual architecture exploration.
SamyakJhaveri/loam/seed/.claude/skills/grill-with-docs Skill
0★ repoGrilling session that challenges your plan against the existing domain model, sharpens terminology, and updates documentation (DOMAIN.md, ADRs) inline as decisions crystallise. Use when user wants to stress-test a plan against their project's language and documented decisions. NOT for implementation review (use /multi-review) or architecture analysis (use /improve-codebase-architecture).
SamyakJhaveri/loam/seed/.claude/skills/improve-codebase-architecture Skill
0★ repoFind deepening opportunities in a codebase, informed by the domain language in DOMAIN.md and the decisions in docs/adr/. Use when the user wants to improve architecture, find refactoring opportunities, consolidate tightly-coupled modules, or make a codebase more testable and AI-navigable. NOT for single-file refactors, performance optimization, or bug fixes (use /fix-bug).
SamyakJhaveri/loam/seed/.claude/skills/scaffold-context Skill
0★ repoAuthor a CONTEXT.md routing file for a specified project subdirectory using the canonical CONTEXT.md anatomy (.claude/rules/context-md-anatomy.md). Use when adding a new high-traffic area whose routing logic the root CLAUDE.md cannot describe economically. Skip for trivial directories. NOT for: editing existing CONTEXT.md files, authoring CLAUDE.md, or documenting project architecture — only new L1 routing files.
SamyakJhaveri/loam/seed/.claude/skills/techdebt Skill
0★ repoScan codebase for tech debt: duplicated logic, dead code, magic numbers. Use at end of session before final commit for a broad codebase sweep. NOT for reviewing only recently changed code (use /simplify).
ats4321/claude-engineering-skills/skills/architecture-analysis Skill
0★ repoAnalyze and design system structure. Auto-load when asked to "evaluate this architecture", "design the module structure", "where should this boundary go", "is this error fatal", "trace the data flow", "should we split this module", or when reviewing how phases/layers of a system connect. Covers contracts-as-seams at phase boundaries, fatal-vs-recoverable error taxonomies, trust boundaries and where validation must live, end-to-end data-flow tracing, criteria for justified module splits, and a runbook for evaluating an unfamiliar architecture without over-engineering.
ats4321/claude-engineering-skills/skills/change-control Skill
0★ repoSafe-change discipline for editing ANY repository. Auto-load before editing files, fixing bugs, refactoring, committing, pushing, publishing, deleting, uploading, or running any irreversible command. Covers read-before-edit, smallest-correct-diff at the root cause (not the symptom), checking every caller before changing a shared function, verifying flag/command behavior (e.g. DRY_RUN) before destructive actions, commit hygiene, when to branch, and never presenting unverified work as done. Trigger keywords: edit, fix, refactor, commit, push, publish, release, delete, migrate, rename, deploy, "make a change", "apply the fix".
ats4321/claude-engineering-skills/skills/codebase-onboarding Skill
0★ repoSystematic principal-engineer onboarding into any unknown repository. Auto-load when asked to "understand this codebase", "get up to speed", "explore this repo", "how does this project work", "map the structure", or before making the first change in an unfamiliar repository. Covers discovery order (README → build config → entry points → tests → CI → git history), what to read first, how to map module structure, and how to distinguish deliberate convention from historical accident. Repo-agnostic: works for Python, Node, Rust, monorepos.
ats4321/claude-engineering-skills/skills/configuration-management Skill
0★ repoDiscovering and designing configuration in any repo. Load when asking "where does this value come from?", when adding a new setting, when wiring env vars / .env files / CLI flags / config files, when deciding setting-vs-constant, or when handling secrets (tokens, webhook secrets, API keys). Covers config-source discovery and precedence, immutable/frozen settings objects, startup validation, .env.example documentation, secrets hygiene (never commit .env), and when NOT to add config at all.
ats4321/claude-engineering-skills/skills/engineering-minimalism Skill
0★ repoThe doctrine skill — decide whether code should exist at all before deciding how to write it. Auto-load when adding features, dependencies, abstractions, config, tests, or security measures; when reviewing a design or PR for over-engineering; when tempted by a framework, wrapper, or "for later" scaffolding; or when the user asks to simplify, trim, or justify complexity. Provides the decision ladder (real need → codebase → stdlib → platform → existing dep → one line → minimal new code), deletion-over-addition, proportionality sizing, guardrail-vs-boundary honesty, and the small-sharp-tool architecture — plus the explicit list of places minimalism must NOT be applied.
ats4321/claude-engineering-skills/skills/llm-system-design Skill
0★ repoDecide WHETHER and WHERE an LLM belongs in a system, and shape the pipeline around it. Auto-load when designing, architecting, or proposing any LLM-backed feature or system — "should we use an LLM for this", choosing between local (Ollama) and hosted (Anthropic/OpenAI) models, selecting model size, choosing pipeline shape (single call, chain, fan-out, agent loop), budgeting cost/latency per call, or designing the failure story for an AI feature. NOT for debugging malformed outputs or parsing (that is llm-integration-reliability) and NOT for building the agent loop itself (that is agent-engineering).
ats4321/claude-engineering-skills/skills/refactoring-playbook Skill
0★ repoSafely restructure working code without changing its behavior. Auto-load when asked to refactor, restructure, extract, rename, move, split, consolidate, modernize, or "clean up" existing code; when migrating from one implementation/pattern to another (strangler-fig); when a change requires reshaping code first; or when deciding whether a refactor is worth doing at all. NOT for fixing bugs (debugging-playbook then change-control), NOT for deciding what code should exist (engineering-minimalism), and NOT for general change safety (change-control governs the diff; this skill governs the behavior-preservation strategy).
ats4321/claude-engineering-skills/skills/system-design Skill
0★ repoDesign NEW systems and components from first principles. Auto-load when asked to "design", "architect", or "propose" a new system, service, feature, or component; when writing a design doc or ADR; when comparing candidate architectures; when estimating capacity or load; or when planning how to deliver a design incrementally. NOT for analyzing an existing system's structure (architecture-analysis), NOT for designing the API surface itself (api-and-interface-design), and NOT for LLM-specific design decisions (llm-system-design).
yousefkadah/laravel-ai-kit/skills/bootstrap Skill
0★ repoUse ONCE right after installing this kit into a project (or when onboarding an agent to an unfamiliar repo). Scans the codebase and fills in .ai/context.md + .ai/architecture.md, replaces the «placeholders», and drafts project-specific skills — turning the blank template into a project-aware guide.
PSB1234/Fabric-Minecraft-Skill-1.20.1/fabric-config Skill
no license0★ repoReference for mod configuration in Fabric Minecraft 1.20.1 — using Cloth Config API for GUI-based config screens, and plain TOML/JSON files with a custom serializer. Use this skill whenever the user asks about mod config, settings file, config screen, Cloth Config, ClothConfigScreen, AutoConfig, TOML config, JSON config, persistent mod settings, or config GUI. Triggers on: "mod config", "config file", "config screen", "Cloth Config", "ClothConfigScreen", "AutoConfig", "TOML", "settings screen", "config serializer", "persistent settings", or "mod options". Part of the fabric-mc-modding skill family.
kulapoo/opencode-agent-harness/.opencode/skills/init-tech-declaration Skill
0★ repoDetects the project's tech and writes a `.opencode/harness/rules/tech.md` router so future sessions skip runtime detection and lazy-load the right conventions. Run once when adopting the harness, and again whenever the tech changes. Triggers on "init tech", "declare my tech", "refresh tech declaration".
Refactor comments that are code smells
ceilidhboy/skills/skills/refactor-comments-that-are-code-smells Skill
0★ repoScans code for comments that explain what code does and refactors them using the action pattern or extracting to methods. Use after writing or editing any code.
Cocos creator scene less test bootstrap
ChrisLamDev/cocos-creator-debug-skills/skills/cocos-creator-scene-less-test-bootstrap Skill
0★ repoUse when creating a runtime test without a .scene file in Cocos Creator 3.x. triggers: - scene less - no scene - test without scene
cat2000/domain-knowledge-ops/.cursor/skills/setup-domain-ops Skill
0★ repoOne-time setup: credentials, company/business intro → industry module template (confirm), Confluence space URL(s), Jira board id(s). Writes team-roots.json as v3 (libraries + teams mounting them). See docs/TEAM_ROOTS_V3.md.
DejusDevspace/my-skills/project-onboarding Skill
no license0★ repoRepository onboarding workflow for agents before contributing code. Use when an agent is asked to familiarize itself with a project, understand an existing repository, identify implementation principles and conventions, map architecture, data flow, naming, caching, testing, build/deploy patterns, or prepare to work on future features with minimal mistakes.
huguryildiz/ALETHEIA/skills/skill-library-generator Skill
0★ repoUse when adopting the Aletheia discipline in a new or existing research repository — discovering the project's actual conventions, interviewing the maintainer to fill the config bindings, extracting the project's own recurring failure modes into project-local skills, and scaffolding the record surfaces (decision log, build log, evidence dirs) after approval. Trigger phrases: "adopt aletheia here", "set up the research discipline", "generate skills for this repo", "bind the pack to this project", "run the adoption interview".
IcodeNet/agent-skills/skills/improve-codebase-architecture Skill
0★ repoFind deepening opportunities in a codebase, informed by the domain language in CONTEXT.md and the decisions in docs/adr/. Use when the user wants to improve architecture, find refactoring opportunities, consolidate tightly-coupled modules, or make a codebase more testable and AI-navigable.
The same file is in 11 other repositories
IcodeNet/agent-skills/skills/setup-engineering-skills Skill
0★ repoSets up an `## Agent skills` block in AGENTS.md/CLAUDE.md and `docs/agents/` so the engineering skills know this repo's issue tracker (GitHub or local markdown), triage label vocabulary, and domain doc layout. Run before first use of `to-issues`, `to-prd`, `triage`, `diagnose`, `tdd`, `improve-codebase-architecture`, or `zoom-out` — or if those skills appear to be missing context about the issue tracker, triage labels, or domain docs.
IcodeNet/agent-skills/skills/system-design Skill
0★ repoDesigns services, APIs, and data models from requirements. Use when designing a new service, API, or data model, or when doing system design for a greenfield or major new capability. Do not use for a small local design choice inside an existing module (use architecture).
IcodeNet/agent-skills/skills/tech-debt Skill
0★ repoScans and prioritizes tech debt and code-health cleanups. Use when asking what to clean up, assessing code health, planning refactors, or prioritizing debt. Do not use for a single design decision (architecture) or for implementing a chosen deepening refactor without a scan.
shinzoxD/knackbox/skills/coding/api-design Skill
0★ repoDesign and review HTTP, REST, GraphQL, and internal service APIs with clear contracts, errors, pagination, security, and evolution rules. Use whenever the user is planning endpoints, reviewing an API specification, naming resources, defining request or response shapes, or asking how services should communicate, even if they only share a rough route list.
shinzoxD/knackbox/skills/coding/graphql-schema-design Skill
0★ repoDesign and review GraphQL schemas for authz, pagination, N+1 risk, nullability, and evolution. Use whenever the user works on GraphQL types, resolvers, connections, deprecations, query cost, or schema reviews — even when they only paste an SDL snippet.
shinzoxD/knackbox/skills/coding/helm-chart-scaffolding Skill
0★ repoScaffold and review Helm charts with safe defaults for values, templates, and hooks. Use whenever the user asks for a Helm chart, Chart.yaml, values.yaml production review, Helm upgrade hooks, or packaging a service for Helm install/upgrade.
shinzoxD/knackbox/skills/coding/refactor-planning Skill
0★ repoPlan safe refactors that preserve behavior — seams, steps, tests, and rollback. Use whenever the user asks how to refactor, untangle a module, reduce complexity, extract a service, or restructure code without changing product behavior — even for "clean this up" requests with a large file.
shinzoxD/knackbox/skills/documents/architecture-decision-records Skill
0★ repoWrite concise Architecture Decision Records that capture context, decision, and consequences. Use whenever the user asks for an ADR, architecture decision log entry, decision record, or why we chose technology X over Y in a durable doc — even for small stack choices.
shinzoxD/knackbox/skills/documents/onboarding-guides Skill
0★ repoWrite new-hire and contributor onboarding guides that get people productive fast. Use whenever the user asks for an onboarding doc, starter guide for engineers, contributor getting-started, first-week plan, or "how do we onboard" handbook section.
knownasnaffy/prompthound/dataset/case_00001 Skill
no license0★ repoInstall and operate local NVIDIA Parakeet ASR for OpenClaw with an OpenAI-compatible transcription API on Ubuntu/Linux and macOS (Intel/Apple Silicon). Use when the user wants private/local speech-to-text, voice transcription setup, ASR troubleshooting, or OpenClaw voice stack configuration with Parakeet (and optional Whisper fallback).
knownasnaffy/prompthound/dataset/case_00047 Skill
no license0★ repoUse nanobanana-plus over HTTP to generate images with per-call model switching and aspect ratio control. Best for OpenClaw / 小龙虾 users who have a running nanobanana-plus service. Run init for a guided setup, then use: node {baseDir}/scripts/nanobanana-plus.mjs generate --prompt 'desc' --filename 'out.png' [--aspect-ratio 16:9] [--model gemini-3.1-flash-image-preview]. Also supports check and models.
knownasnaffy/prompthound/dataset/case_00067 Skill
no license0★ repoRun Cursor Agent CLI for coding tasks — writing, editing, refactoring, reviewing, or planning code — without spending OpenClaw API credits. Use when the user asks to write/fix/refactor/review code, a coding task would otherwise be done inline with Sonnet/Haiku, the user says "do this in cursor" or "use cursor for this", or any substantial file-editing task in a known repo. NOT for conversational questions about code (answer inline) or tiny one-liners that don't warrant a subprocess.
knownasnaffy/prompthound/dataset/case_00079 Skill
no license0★ repoSync AI coding assistant config files from a single source. Generates CLAUDE.md, AGENTS.md, GEMINI.md, .cursor/rules/, .github/copilot-instructions.md, .windsurf/rules/, .clinerules/, CONVENTIONS.md, .kiro/steering/, .amazonq/rules/, .goosehints, .trae/rules/ from your canonical rules. ACTIVAR cuando el usuario dice 'sync configs', 'generate ai rules', 'configurar para otros tools', 'unificar instrucciones', 'config sync'.
knownasnaffy/prompthound/dataset/case_00151 Skill
no license0★ repo以用户可执行的工作流方式安装、配置和排查 HiLight OpenClaw 插件。用户想把 OpenClaw 连接到 HiLight、安装 `@art_style666/hi-light` 插件、把 `channels["hi-light"]` 写入 OpenClaw 配置、更新 HiLight API Key 或 WebSocket 地址、或排查 HiLight 无法连接时使用。 Install, configure, and troubleshoot the HiLight OpenClaw plugin as a user-facing setup workflow. Use when a user wants to connect OpenClaw to HiLight, install `@art_style666/hi-light`, write `channels["hi-light"]` into OpenClaw config, update the HiLight API key or websocket URL, or diagnose why the HiLight channel does not connect.
knownasnaffy/prompthound/dataset/case_00176 Skill
no license0★ repoUse when creating Unity game scenes and prototypes, building level designs, or automating Unity Editor workflows for terrain, lighting, environment setup, and player controllers. Use for post-apocalyptic scenes, fantasy forests, dungeon crawlers, and other game environments that need rapid prototyping.
knownasnaffy/prompthound/dataset/case_00270 Skill
no license0★ repoBootstrap a new OpenClaw agent's memory system in one command. Sets up MEMORY.md, daily memory files, HEARTBEAT.md, and USER.md by asking 3 quick questions. Use when a user is setting up OpenClaw for the first time, when memory files are missing, or when asked to "set up my memory system", "initialize my agent", "bootstrap my agent", or "run the memory wizard". Solves the #1 OpenClaw pain point — agents waking up with no context about who they're serving.
knownasnaffy/prompthound/dataset/case_00301 Skill
no license0★ repo以用户可执行的工作流方式安装、配置和排查 HiLight OpenClaw 插件。用户想把 OpenClaw 连接到 HiLight、安装 `@art_style666/hi-light` 插件、把 `channels["hi-light"]` 写入 OpenClaw 配置、更新 HiLight API Key 或 WebSocket 地址、或排查 HiLight 无法连接时使用。 Install, configure, and troubleshoot the HiLight OpenClaw plugin as a user-facing setup workflow. Use when a user wants to connect OpenClaw to HiLight, install `@art_style666/hi-light`, write `channels["hi-light"]` into OpenClaw config, update the HiLight API key or websocket URL, or diagnose why the HiLight channel does not connect.
knownasnaffy/prompthound/dataset/case_00323 Skill
no license0★ repoConfigure a MoltBot agent to participate in a Telegram group chat. Automates adding the group to the allowlist, setting mention patterns, and configuring sender permissions — all via a single gateway config patch. Use when the user wants to set up their bot in a Telegram group, enable cross-bot communication, or configure group mention gating.
knownasnaffy/prompthound/dataset/case_00329 Skill
no license0★ repoMigrate settings from EasyClaw into OpenClaw by locating EasyClaw desktop/runtime config files, comparing them with ~/.openclaw/openclaw.json, and safely generating or applying a selective merge. Use when a user mentions EasyClaw, wants to copy settings/configs from EasyClaw into OpenClaw, asks where EasyClaw stores its config, or needs a migration report or backup-aware merge.
knownasnaffy/prompthound/dataset/case_00373 Skill
no license0★ repo华为云Config配置审计服务的专业助手,提供全局资源查询、资源配置查询、资源Schema查询、SQL条件查询以及合规规则查询能力。当用户需要查询华为云资源、进行资源审计、检查配置合规性、盘点云上资产、分析资源关系、查询特定资源状态、验证资源配置、或需要通过SQL进行复杂资源查询时,必须使用此技能。即使未明确提及"Config"或"配置审计",只要涉及华为云资源查询、资源管理、合规检查等场景,都应触发此技能。
knownasnaffy/prompthound/dataset/case_00377 Skill
no license0★ repoInstall and configure headless Chrome for OpenClaw browser tool in environments without root/sudo access (cloud containers, VPS, sandboxed hosts). Use when: (1) browser tool fails with No supported browser found, (2) Chrome crashes with Page crashed or missing library errors, (3) setting up browser automation on a fresh server or container, (4) user asks to install or fix the browser for OpenClaw. NOT for macOS/Windows desktops with Chrome already installed, or Chrome extension relay setup.
knownasnaffy/prompthound/dataset/case_00465 Skill
no license0★ repoRun Cursor Agent CLI for coding tasks — writing, editing, refactoring, reviewing, or planning code — without spending OpenClaw API credits. Use when the user asks to write/fix/refactor/review code, a coding task would otherwise be done inline with Sonnet/Haiku, the user says "do this in cursor" or "use cursor for this", or any substantial file-editing task in a known repo. NOT for conversational questions about code (answer inline) or tiny one-liners that don't warrant a subprocess.
knownasnaffy/prompthound/dataset/case_00528 Skill
no license0★ repoNavigate and use OpenClaw documentation efficiently with cached doc fetch/search tooling, sitemap/category routing, and config snippet lookup. Use when answering OpenClaw setup/config/troubleshooting questions, checking docs updates, or extracting concrete configuration examples from docs.
knownasnaffy/prompthound/dataset/case_00621 Skill
no license0★ repoReference tool for devtools — covers intro, quickstart, patterns and more. Quick lookup for Windows Setup concepts, best practices, and implementation patterns.
knownasnaffy/prompthound/dataset/case_00694 Skill
no license0★ repoUse this skill to initialize or update a multi-repo workspace created from dev-config-template. Invoke whenever the user wants to: set up a fresh workspace clone, fill in CLAUDE.md placeholders, clone sub-repos, configure dev environments, or sync workspace files with the latest template. Triggers on: "init workspace", "initialize", "set up my project", "configure workspace", "update template", "sync template", "update workspace", dev-config-template setup questions, or Chinese: 初始化工作区, 配置工作区, 更新工作区模板, 模板有更新吗. Do NOT trigger for: project scheduling, weekly reports, or individual git operations.
knownasnaffy/prompthound/dataset/case_00722 Skill
no license0★ repoReference tool for devtools — covers intro, quickstart, patterns and more. Quick lookup for Windows Setup concepts, best practices, and implementation patterns.
knownasnaffy/prompthound/dataset/case_00769 Skill
no license0★ repoDiagnose OpenClaw setup in one command. Auto-detects issues with Node, npm, gateway, config, and workspace. Quick mode (~10s) or Full mode. Multi-language. Config and workspace checks report file existence only — never reads contents. No credentials accessed without approval. Homepage: https://clawhub.ai/skills/setup-doctor
knownasnaffy/prompthound/dataset/case_00793 Skill
no license0★ repoSet up the full OpenClaw agent memory system with 3-tier memory (HOT/WARM/COLD), daily logs, semantic search (QMD), and lossless context management (Lossless Claw). Use when onboarding a new agent, setting up memory for a fresh OpenClaw instance, or when asked to install the memory system on a new agent. Triggers on "set up memory", "install memory system", "onboard new agent memory", "memory setup", "agent onboarding", "configure agent memory", "add memory to my agent", "how do I set up memory", "initialize memory", "memory system for OpenClaw".
knownasnaffy/prompthound/dataset/case_00798 Skill
no license0★ repoOpenClaw HowTo skill — the accurate, reliable guide for using OpenClaw. Provides authoritative answers on OpenClaw features, configurations, CLI commands, troubleshooting, and best practices. Supports dual modes: 1) Web Search mode (using OpenClaw built-in or user-configured web_search tool for online queries); 2) Local mode (obtaining information via CLI/config/knowledge base). Will remember user configuration after first use and no longer ask repeatedly.
knownasnaffy/prompthound/dataset/case_00830 Skill
no license0★ repoManage Control D DNS filtering service via API. Use for DNS profile management, device configuration, custom blocking rules, service filtering, analytics settings, and network diagnostics. Triggers when user mentions Control D, DNS filtering, DNS blocking, device DNS setup, or managing DNS profiles.
knownasnaffy/prompthound/dataset/case_00837 Skill
no license0★ repoAutomatically onboard new Telegram groups so the bot responds immediately — with project setup. Use when: (1) a user says they're creating a new Telegram group with the bot, (2) the bot was added to a new group but doesn't respond, (3) someone reports the bot ignores messages in a group. Solves the groupPolicy=allowlist chicken-and-egg problem by watching gateway logs for dropped messages, extracting the chat_id, adding it to config with requireMention=false, creating a project for the group, and hot-reloading. One smooth flow.