Folken2
- 43 skills
- 0 followers
- 5 hours ago last updated
- ▌ Style Learning Loop 2 · folken2How to learn and refine the user's writing style from passages they keep after agent edits, and when to consolidate fingerprints into a tight rulebook
- ▌ Voice Matching 2 · folken2How to draft email replies and new messages that sound like the user, using the style memory
- ▌ Style Learning Loop 3 · folken2How to learn and refine the user's writing style from their sent emails, and when to consolidate fingerprints into a tight rulebook
- ▌ Getting Started 2 · folken2How this Hermes profile is put together — SOUL.md, config.yaml, and the skills directory — and how to add a skill. Read when asked what this agent can do, when someone wants to extend it with a new capability, or as a worked example of the SKILL.md format.
- ▌ Nuvel · folken2Use when the user wants to build a Google ADK agent — scaffolding a new agent, adding tools/skills/prompts to one, or asking how to structure an ADK project. Triggers on phrases like "create an ADK agent", "build a Google agent", "scaffold an agent", "make an agent that does X", "I need an ADK skeleton", or any task involving generated-agents/, LlmAgent, SkillToolset, or the ADK framework. Also use whenever the user mentions `nuvel`, the `nuvel` CLI, or asks about agent architecture patterns / callbacks / HITL / streaming / ADK prompt engineering — nuvel ships the canonical knowledge skills for those topics. Lean toward triggering — if the task touches Google ADK at all, this skill is in scope.
- ▌ Buzz · folken2Building and operating Buzz agents — env-var configuration, the ACP stdio adapter, Nostr identity and event signing, joining a NIP-29 relay group, and rendering skills into a Buzz persona. Read when scaffolding with `--framework buzz`, when an agent needs to speak in a Buzz group, when deciding how a Buzz agent should be configured or deployed, or when a relay connection or signature is being rejected.
- ▌ Hermes · folken2Building and operating Hermes agents — the SOUL.md / config.yaml / skills profile shape, model and turn-budget config, installing a profile into a Hermes home, and the Telegram gateway with its DM policy. Read when scaffolding with `--framework hermes`, when a Hermes bot needs to answer on Telegram, when deciding what belongs in SOUL.md versus config.yaml, or when a profile isn't loading.
- ▌ Adk Streaming · folken2 bundleBuilding voice and video agents with Gemini Live API — bidirectional streaming via WebSocket, LiveRequestQueue, run_live(), RunConfig, and audio/video patterns. Load when the user wants a voice agent, live agent, or real-time audio/video capabilities.
- ▌ Adk Tool Creation · folken2 bundleBuilding Google ADK function tools — ToolContext usage, proper signatures, error handling, return formats, type hints, and async patterns. Load this skill when creating custom tools for an agent.
- ▌ Adk Agent Patterns · folken2 bundleAgent architecture patterns for Google ADK 2.0 — when to reach for a single LlmAgent, a Workflow graph (new default for multi-step orchestration), or the shortcut classes SequentialAgent / LoopAgent / ParallelAgent. Load this skill when deciding the agent's top-level shape.
- ▌ Adk Callbacks Hitl · folken2 bundleADK callback signatures, HITL approval gates, and state management — exact parameter names, before/after hooks, EventActions.state_delta, and defensive HITL patterns. Load this skill when adding callbacks or approval gates to an agent.
- ▌ Adk Cron Isolation · folken2Blast-radius control for scheduled (cron) runs of an ADK agent — scoped secrets, the headless tool-approval policy, and HITL-gated job creation. Read when an agent runs jobs on a schedule with no human present, when a cron job needs credentials but shouldn't see every env var, when a scheduled tool call is being denied unexpectedly, or when tuning NUVEL_CRON_HEADLESS_POLICY / NUVEL_CRON_SCOPE_SECRETS / NUVEL_CRON_HITL_CREATE.
- ▌ Adk Skill Creation · folken2 bundleCreating valid SKILL.md files following the agentskills.io specification — frontmatter, instructions, references directory, progressive disclosure (L1/L2/L3), and SkillToolset wiring in agent.py. Load this skill when generating domain skills for an agent.
- ▌ Adk Task Delegation · folken2 bundleDelegate work to sub-agents with the ADK 2.0 Task API — `mode='task'`, `mode='single_turn'`, `mode='chat'` on `LlmAgent`, the auto-attached `finish_task` tool, and typed contracts via `input_schema` / `output_schema`. Load this skill when one agent needs to hand a bounded unit of work to another and get a validated result back, or when migrating off SequentialAgent / ParallelAgent / LoopAgent.
- ▌ Adk Workflow Graphs · folken2 bundleBuild graph-based agents with ADK 2.0 `Workflow` — declare nodes and edges, route conditionally, fan-out/fan-in in parallel, run dynamic nodes at runtime, and add human-in-the-loop revision cycles. Load this skill whenever the agent needs anything beyond a strictly linear or trivially parallel pipeline.
- ▌ Adk Prompt Engineering · folken2 bundleWriting effective ADK system prompts — dynamic instructions with InstructionProvider, state placeholders, context injection, tone/style sections. Load this skill when creating or improving an agent's system prompt.
- ▌ Adk Composio Tool Router · folken2 bundleThe Composio Tool Router for ADK agents — the --with-composio scaffold flag, per-user sessions via composio.create(user_id=...), the McpToolset wiring, multi-tenant patterns, and when not to reach for it. Read when an ADK agent needs broad integration coverage (Gmail, GitHub, Slack, Notion, Calendar, Linear, etc.) without authoring per-service tools, when the user wants ~1000 toolkits behind one endpoint, or when adding multi-tenant credential isolation to an agent.
- ▌ Adk Org Memory Retrieval · folken2 bundleScope-aware hierarchical memory for ADK agents via OrgMemoryService — wiring it through ADK's service registry with three env vars, plus how retrieval works (hybrid RRF fusion of keyword and vector arms, tier boost, floor-gated boost cascade, autocut, a zero-LLM knowledge graph, relational recall, and answer synthesis). Read when an agent needs memory shared across users/teams/an org rather than per-session, when wiring NUVEL_ORG_MEMORY_URI / NUVEL_ORG_MEMORY_DSN / NUVEL_ORG_GRAPH_PATH, when tuning retrieval quality, or when deciding whether org memory is warranted at all.
- ▌ Adk Long Horizon Sessions · folken2Surviving long ADK runs — resumability so an interrupted invocation continues instead of restarting, and event compaction so a long session doesn't exhaust the context window. Read when an agent runs for hours, when a session's history is growing past the context window, when deciding whether RESUMABILITY should stay on for a stateless deployment, or when tuning COMPACTION_INTERVAL / COMPACTION_OVERLAP / COMPACTION_RETENTION. Pairs with adk-long-horizon-guardrails (which stops runaway runs) and adk-prompt-engineering (the cache-stable prompt tiers, a separate but adjacent long-horizon cost lever).
- ▌ Adk Skill Design Patterns · folken2 bundleFive canonical skill design patterns for ADK agents: Tool Wrapper, Generator, Reviewer, Inversion, and Pipeline. Load this skill during Step 2 (Design) to recommend which patterns fit the user's agent, and during Step 4 (Generate) to get skeleton templates for each pattern.
- ▌ Adk Long Horizon Guardrails · folken2 bundleHalt guards and command-safety guardrails for ADK agents — the shared halt latch, NoProgressGuard, RepeatedFailureGuard, the structural argv-level shell command classifier, and exfil_guard. Read when an agent will run unattended or for a long time, when a run needs a backstop against response loops or a tool retried identically forever, when an agent can execute shell commands, or when tool arguments might carry secrets. Also read when tuning EXFIL_GUARD_STRICT or debugging a "[halted: ...]" response.
- ▌ Adk Memory Self Improvement · folken2 bundleEXPERIMENTAL, opt-in memory self-improvement for long-lived ADK agents — relevance-conditioned preload, the periodic consolidation "dream" pass, the after-turn judge fork, fire-and-forget sibling runs, per-session throttling, and the skill curator that proposes SKILL.md files. Read before enabling any NUVEL_MEMORY_* or NUVEL_SKILL* flag, when budgeting the extra LLM calls these add per turn, or when an agent should improve its own memory and skills over months. Every fork-adding feature here defaults off (preload is the on-by-default exception) and none should reach production without eval coverage.
- ▌ Claude Sdk Hooks · folken2 bundleHooks in the Claude Agent SDK — PreToolUse, PostToolUse, UserPromptSubmit, Stop event types, the HookMatcher signature, hook function signatures, blocking vs auditing patterns, and when hooks beat permission callbacks. Read when adding deterministic guards around tool calls, when implementing audit logging, when injecting context dynamically, or when can_use_tool isn't enough.
- ▌ Deck Outlining · folken2Pattern for turning a brief into a coherent deck outline — intent detection, section ratios, draft headings, expansion
- ▌ Deck Structure · folken2When and how to suggest reordering, splitting, or trimming a deck — agenda placement, CTA placement, problem-solution arc, methodology-results arc
- ▌
- ▌ Slide Tightening · folken2Rubric for sharpening the active slide — parallelism, bullet count, length, title strength, notes-vs-bullets separation
- ▌ Rewrite Rubric · folken2How to rewrite a selected passage per the user's instruction without silently expanding scope, breaking length, or losing quoted material
- ▌ Voice Matching · folken2How to draft new Word sections and extensions that sound like the user, using the style memory and the document context
- ▌ Claude Sdk Deployment · folken2Deploying Claude Agent SDK agents to production — wrapping the SDK in FastAPI with SSE streaming, session resume, cost tracking via ResultMessage, Docker/Railway/Fly config, ANTHROPIC_API_KEY handling, and bearer-auth patterns. Read when going from "it works locally" to "it's running on a server", when designing the request/response contract for an agent endpoint, or when investigating why deployments are slower or more expensive than expected.
- ▌ Style Learning Loop · folken2How to learn the user's deck style from kept slides and when to consolidate fingerprints into a tight rulebook
- ▌ Claude Sdk Permissions · folken2The Claude Agent SDK permission model — allowed_tools, disallowed_tools, the six permission_mode values, and the can_use_tool callback. Read when designing what an agent is allowed to do, when production deploys are surprising users with permission prompts, when you want dynamic per-call gating, or when locking down a deploy that runs unattended.
- ▌ Claude Sdk System Prompt · folken2System prompts in the Claude Agent SDK — string vs file vs preset, the claude_code preset and its append field, dynamic prompts via per-turn rebuilding, and how setting_sources composes user/project/local instructions. Read when designing the agent's voice and operating principles, when migrating an existing prompt into the SDK shape, or when choosing between a static and a dynamic prompt.
- ▌ Claude Sdk Tool Creation · folken2 bundleBuilding custom tools for Claude Agent SDK agents — the @tool decorator, schema dicts, return shape, error returns, and the mcp__server__tool naming gotcha. Read when adding a new tool to a Claude Agent SDK project, when you need to expose Python functions to Claude, or when designing tool boundaries.
- ▌ Email Coaching · folken2Rubric for giving honest, specific, voice-aware feedback on the user's draft emails
- ▌ Claude Sdk MCP Integration · folken2Wiring external MCP servers into a Claude Agent SDK agent — stdio and HTTP transports, dict format, when to use SDK MCP vs external, naming conventions, and common pre-built servers (filesystem, fetch, sequential-thinking). Read when you need to give Claude access to a service via an existing MCP server, when choosing between writing your own tools and consuming an external one, or when debugging "tool not found" errors.
- ▌ Managed Agents Tools · folken2The three kinds of tools in Anthropic Managed Agents — the prebuilt agent toolset (`agent_toolset_20260401`), MCP toolsets with vault-backed credentials, and custom tools handled host-side. Read when adding a new capability to your agent, when deciding where to put a third-party API call, when a third-party service has a hosted MCP server, or when an integration needs an API key.
- ▌ Managed Agents Events · folken2The event stream model for Anthropic Managed Agents — stream-first ordering, the idle-break gate, lossless reconnect via consolidation, custom-tool round-trip, and the most-common event types. Read when implementing or debugging the orchestrator loop, when sessions hang or terminate prematurely, or when reconnecting a dropped SSE stream.
- ▌ Managed Agents Overview · folken2The mandatory flow for Anthropic Managed Agents — create a versioned agent once, reference it from sessions on every run. Read first when starting a Managed Agents project, when migrating from a different Anthropic surface (Claude API, Claude Agent SDK), or when confused about why session.create() rejects model/system/tools fields.
- ▌ Getting Started · folken2How this agent's skills work and how to add one. Read when asked what you can do, when someone wants to extend you with a new capability, or as a worked example of the SKILL.md format.
- ▌ Managed Agents Deployment · folken2Deploying a Managed Agents project to production — wrapping the SDK in FastAPI with SSE, the env vars that have to be set, agent-version pinning, bearer auth, cost tracking via span events, and Docker/Railway/Fly. Read when going from "setup.py worked locally" to "it's running on a server", when designing the request/response contract, when investigating why deployments lose state, or when planning rollouts of agent updates.
- ▌ Managed Agents Skills Memory · folken2Skills and memory stores in Anthropic Managed Agents — Anthropic prebuilt skills (xlsx/docx/pptx/pdf), custom skills via the Skills API, and memory stores for cross-session persistence with versioning and audit. Read when adding domain expertise the agent should auto-load, when the agent needs persistent state across sessions, or when distinguishing memory from session resources.
- ▌ Outlook Search Via Composio · folken2How to search the user's Outlook mailbox effectively using Composio's OUTLOOK_* tools, including query planning and result ranking