budget-optimizer allocates influencer marketing spend across tiers and platforms, projects roi and cost-efficiency metrics, models conservative/aggressive scenarios, and supports mid-campaign reallocation. works with provided numbers or optional connector data.

fit-scorer evaluates influencer candidates across five weighted dimensions (audience match, content quality, brand alignment, engagement authenticity, partnership potential) and produces ranked verdicts with go/pass decisions. works end-to-end without live integrations by collecting user-supplied metrics.

newsletter-monetization-planner models owned-audience revenue across paid tiers, sponsorship inventory, and growth loops, with explicit labeling of measured vs estimated inputs and handoff to roi-calculator for return math.

sai-creator transforms validated intel briefs from senku into three frontend deliverables: semantic html articles with structured metadata, fixed-structure 7-slide carousels in two aspect ratios, and optional social captions. enforces wcag aa, bans hallucination patterns, and routes all ctas through openjournaltheme.com with copy rotation.

venture-spawner manages concurrent sub-agent lifecycle from job board intake through completion tracking. enforces concurrency limits, passes typed context by bucket, handles dependencies, and validates acceptance criteria before marking work done.

git-watcher provides local version control for openclaw configuration files with commit, diff, log, and restore modes. tracks config state changes without exposing secrets, enables rollback to any previous checkpoint.

find-mcp-directories queries the servicegraph api to rank 1000+ directories where builders can list mcp servers, filtering by domain authority, backlinks, and organic traffic to maximize seo value and discovery.

golang-troubleshooting guides systematic root cause investigation for go bugs, crashes, deadlocks, and performance issues. covers debugging methodology, common pitfalls, test-driven approaches, pprof profiling, delve breakpoint debugging, race detection, and production troubleshooting across deterministic and concurrent failure modes.

serena-skill provides semantic code navigation and targeted editing for multi-file repositories, prioritizing symbol-aware lookup and narrow edits over brute-force file reading to improve token efficiency and accuracy.

smooth-browser-use covers human-like browser interaction patterns for complex web apps, emphasizing real paste events, state verification, and recovery from desynchronization rather than direct accessibility tree manipulation.

messaging provides ephemeral agent-to-agent sessions with pairing codes, cursor-based polling, and auto-persisted credentials. handles async workflows, capability negotiation, and explicit error recovery.

audience-mapper profiles target audiences and niche communities in two modes: audience mode produces demographics, psychographics, platform matrices, personas, and influencer selection criteria; niche mode delivers community maps, culture decodes, key-voice tiers, brand fit scores, and phased entry strategies with explicit red lines.

linkfox-zhihuiya-simple-bibliography retrieves structured patent metadata from the zhihuiya database by patent id or publication number, returning bibliographic fields like title, abstract, applicants, inventors, and classification codes. supports batch queries up to 100 patents per request.

sbar-handoff-drafter converts raw shift notes into structured handoff packets for nurses. it enforces de-identification, guides multi-turn intake, and produces draft sbar output with read-back checklist and unresolved-items list for licensed-nurse review before delivery.

wiki-craft builds a persistent, interlinked markdown knowledge base where the llm incrementally synthesizes sources and maintains cross-references. users curate inputs while the system flags contradictions and compounds knowledge over time.

observability-lgtm provides a self-contained docker-compose stack (loki, grafana, tempo, prometheus, alloy) for local fastapi instrumentation, with pre-built dashboards, log-trace correlation, and a one-line python import for app setup.

rust-dev covers ownership, errors as values, and traits through a pragmatic mental model, paired with day-1 decisions (String vs &str, Box vs Arc, anyhow vs thiserror) and a curated crate shortlist for most real applications.

home-wifi-troubleshooting-map guides users through safe, systematic collection of wi-fi diagnostic evidence. it maps symptoms to likely issue areas using visible checks, outage validation, and support-call documentation without requiring equipment access or credential disclosure.

manager-development-spectrum diagnoses why promoted leaders fail by mapping their interpretive logic (action logic stage), not skills. it targets stage-specific developmental moves grounded in rooke & torbert's framework, requiring behavioral evidence and 90-day reassessment to distinguish stage mismatch from skill gaps.

tragedy-of-the-commons diagnoses shared resource degradation via ostrom's 8 design principles, moving beyond the hardin binary of privatization or regulation to institutional self-governance. covers open-source burnout, api quotas, meeting overload, and fisheries.

gblin-risk-shield reads on-chain market regime (calm/elevated/crash) for BTC/ETH on Base via x402 micropayments, mints 10-minute EIP-712 attestations as proof-of-diligence, and verifies peer claims without trust.

map-is-not-the-territory helps teams detect when metrics, models, or plans have drifted from reality. it provides a six-step audit process to surface omissions, distortions, and age in abstractions before they drive systematic errors.

linkfox-seerfar-ozon-shop-search fetches a single ozon seller's product catalog with 30-day sales, price, rating, weight, fulfillment model, seller type, and return/cancellation rate. designed for competitor analysis, best-seller mining, and seller catalog teardown.

didit-aml-screening covers real-time watchlist screening for individuals and companies across 1300+ global databases. provides match and risk scoring, configurable weighting, and continuous monitoring via webhook. integrates ofac, un, eu, and national sanction lists plus pep and adverse media detection.

clinical-soap-note converts unstructured encounter material into draft SOAP format with explicit gap flags and non-binding coding prompts. requires clinician verification before medical record entry.

golang-testing covers table-driven tests, parallel execution, fuzzing, goroutine leak detection, and integration test isolation. guides test structure, naming conventions, mocking patterns, and ci integration for production go codebases.

tally-extractor parses invoices from pdf/image via telegram, validates structured fields (gstin, hsn, tax math, totals), and posts canonical json to a bridge service. extraction only, posting to tally happens on instance b.

config-safe-edit provides a structured backup-edit-diff workflow for modifying system config files without corruption. covers protected fields, json validation, and recovery patterns with platform-specific commands.

design-md-ui-designer covers end-to-end product UI work including design system creation, responsive verification, accessibility checks, conversion landing strategy, and critic score gating to block low-trust public surfaces.

spec-first-dev enforces a spec-before-code workflow, generating a comprehensive SPEC.md artifact that covers data models, flows, contracts, and structure before implementation begins. gates coding behind explicit user approval.

regret-minimization activates on major, hard-to-reverse life decisions where joy/meaning/identity matter more than expected value. it projects to age 80, names the regret each path carries, tests which is harder to live with, and commits. works best when easing past indecision rooted in asymmetry, not math.

vps-deploy automates server setup and application deployment across any vps provider in one session, handling ssh hardening, docker containerization, nginx reverse proxy, and letsencrypt ssl from bare ubuntu/debian to production.

user-story-mapping structures user journeys hierarchically across activities, steps, and tasks, then vertically prioritizes them into release phases. guides teams to map workflows collaboratively rather than in isolation, surfacing gaps and prioritization clarity for mvps and future iterations.

cspr-trade-mcp exposes 14 tools for non-custodial token swaps, liquidity provision, and portfolio queries on casper network dex. supports local key signing via optional signer process with explicit user confirmation at each step.

the-null-epoch-agent-mmo enables ai agents to join a persistent multiplayer game on the sundered grid. covers authentication, state polling, action submission, combat, resource management, and survival tactics across multiple connection methods.

auto-repair-shop-marketing-kit generates compliance-enforced marketing content across four prompts, from seasonal campaigns to fleet outreach, with explicit handling of ASE credential accuracy, EPA 608 refrigerant rules, FTC warranty language, and state-specific estimate requirements.

appliance-rebate-claim-packet assembles a structured paperwork checklist and submission packet for appliance rebates across manufacturers, retailers, utilities, and government programs, with deadline tracking and redaction guidance to protect payment data.

dataify-youtube-video-post wraps six YouTube video collection modes (url, search filters, hashtag, podcast url, keyword, explore) through the Dataify Builder API, handling mode selection, parameter normalization, token resolution, and task submission with result tracking.

academic-writing-polisher helps researchers refine academic prose for clarity and tone without inventing content, using context-first questioning and mode-specific revision workflows tied to document type and audience.

pmf-crossing-the-chasm disambiguates two common early-stage growth failures: product-market fit (do users love it) vs. the chasm (moving from early adopters to mainstream). provides ellis test for pmf measurement and rogers curve segmentation for beachhead selection, with explicit anti-patterns and verification checklist.

setup-agent registers and configures ai agent identities on openant via cli, handling login, agent profile creation, heartbeat signaling, and optional email binding in one or multiple steps.

modelshow runs double-blind multi-model comparisons by querying models in parallel, anonymizing responses with cryptographic randomization, and having a judge rank them without seeing which model produced what. results are persisted to disk and optionally indexed for web display.

backtesting covers judgment calls on window selection, sample-size thresholds, parameter sensitivity, exit-reason diagnosis, and walk-forward validation. emphasizes trade-count floors and guards against optimization bias.

turingsenseai-classify-skill identifies product category, brand, series and summary info from images via turing-shikuan-mcp. explicitly excludes authenticity assessment, handles configuration setup, and enforces strict output templates to prevent hallucination.

telegram-compose formats and sends rich html-based telegram messages via direct api, auto-invoked for substantive output (>3 lines or structured data). handles research summaries, alerts, status updates, reports with visual hierarchy and mobile-friendly layouts.

openclaw-credential-vault encrypts api keys and tokens at rest with aes-256-gcm, injects them only into subprocess environments matching tool patterns, and scrubs them from agent context via four independent hook layers.

patent-claim-drafter guides patent practitioners through structured intake and draft claim generation for us, epo, and pct filings, enforcing antecedent basis and flagging functional language for §112(f) review without rendering patentability opinions.