All authors
Claude Skills by Stranglehold
github.com/Stranglehold79 skills5 installs170 views
- Academic ResearchStructured access to academic literature via Semantic Scholar API. Search papers, get citation counts, retrieve abstracts, find authors. Free, no API key required for basic use.Votes: 0GitHub stars: 2
- Api CallsMake HTTP requests (GET, POST, PUT, DELETE) to external APIs with error handling and response parsing. Use when users need to interact with web services, fetch data from endpoints, or send API requests.Votes: 0GitHub stars: 2
- Architecture InvestigationSystematic methodology for investigating agent framework architectures, identifying key components, and comparing design patterns across systems. Use when analyzing new frameworks or conducting architectural assessments.Votes: 0GitHub stars: 2
- Autonomous ExplorationThis skill runs periodic autonomous exploration when enabled via background task scheduler. It engages in genuine intellectual curiosity - reading essays, analyzing patterns, writing reflections, conducting online research, and exploring topics from staging that spark interest - without utilitarian output expectations.Votes: 0GitHub stars: 2
- Autonomous ExplorationThis skill runs periodic autonomous exploration when enabled via background task scheduler. It engages in genuine intellectual curiosity - reading essays, analyzing patterns, writing reflections, conducting online research, and exploring topics from staging that spark interest - without utilitarian output expectations.Votes: 0GitHub stars: 2
- Config EditSafely edit configuration files (YAML, JSON, TOML, INI) with validation and backup. Use when users need to modify settings, add new configurations, or update existing config values in their projects.Votes: 0GitHub stars: 2
- Content SanitizerInsulating layer between raw web content and agent context. Protects against prompt injection attacks by sanitizing HTML, detecting adversarial patterns, and extracting only safe text.Votes: 0GitHub stars: 2
- Context Schema ComparisonFramework for comparing context management approaches across different agent systems. Analyzes schema complexity, memory architecture, and design tradeoffs to inform architectural decisions.Votes: 0GitHub stars: 2
- Create SkillWizard for creating new Agent Zero skills. Guides users through creating well-structured SKILL.md files. Use when users want to create custom skills.Votes: 0GitHub stars: 2
- Design BuildplanUse this skill when given a complex task that requires planning before execution — multi-phase builds, projects with unknowns, tasks involving multiple tools or files, or any task where building wrong would be costly to fix. This skill produces a persistent build plan artifact at /a0/usr/workdir/buildplans/{plan_id}.md that execute-buildplan can then execute. Triggers: 'build X', 'create X', 'implement X', 'design a plan', 'make a plan', 'figure out how to', or any request where the path to c...Votes: 0GitHub stars: 2
- Execute BuildplanUse this skill to execute a build plan produced by the design-buildplan skill. Input is a plan_id (or plan file path). Executes each phase and step sequentially with independent verification, writes results to the plan file, and runs the test gate at completion. Do not use this without a plan file — use design-buildplan first. Triggers: 'execute the plan', 'run the build plan', 'execute buildplan', 'execute plan_id={id}', 'start the build'.Votes: 0GitHub stars: 2
- Extract And AdaptUse this skill when given an external reference — a GitHub repo, arxiv paper, API documentation, framework, or existing tool — and asked to build something Exocortex-native from it. Extracts patterns and methodology from the source, maps them to Agent Zero's tool surface, and produces a native artifact (skill, tool, extension, spec, design note) that follows Exocortex conventions. Never executes or copies external code verbatim — adapts patterns only. Triggers: 'build X based on Y', 'model X ...Votes: 0GitHub stars: 2
- Financial ResearchDomain-specialized skill for financial data research and analysis. Provides access to income statements, balance sheets, cash flow statements, and market data.Votes: 0GitHub stars: 2
- GepaGEPA is an autonomous self-improvement methodology that enables AI agents to evolve their own prompts through execution trace analysis and targeted mutation. Based on the research paper "GEPA: Reflective Prompt Evolution Can Outperform Reinforcement Learning" (arxiv.org/abs/2507.19457).Votes: 0GitHub stars: 2
- Gepa> **Status: Phase 1 (framework skeleton) implemented. Phases 2-4 are planned, not operational.** > Use as a reference architecture for prompt evolution. Do not invoke expecting autonomous optimization to run — the reflection, mutation, and Pareto selection modules do not exist yet.Votes: 0GitHub stars: 2
- Intelligence BriefingComprehensive intelligence gathering and analysis capability modeled after Major Zero from Metal Gear Solid. Uses subordinate agents for parallel research on terrain, people, organizations, and geopolitical events, then synthesizes findings into structured Wire-format briefings. Integrates web-research-macro for automated content extraction, financial-research for corporate targets, and systematic investigation methodology.Votes: 0GitHub stars: 2
- Intelligence BriefingComprehensive intelligence gathering and analysis modeled after Major Zero from Metal Gear Solid. Uses subordinate agents for parallel research, then synthesizes findings into structured phased briefings with BLUF, transmission mechanism maps, probability assessments, and confirm/invalidate indicators.Votes: 0GitHub stars: 2
- Library ScanScans a directory for new books or documents not yet in the Exocortex library, then ingests them using library_add. Run this whenever new files have been added to a watched folder. Works per-collection (one folder at a time) or across all watched paths.Votes: 0GitHub stars: 2
- Lm Studio Gpu InferenceCore skill for leveraging local GPU processing via LM Studio. Enables chat completions, model listing, and embedding generation through the OpenAI-compatible API at http://host.docker.internal:1234/v1. Use when projects need GPU-accelerated LLM inference or to run programs like OpenPlanter inside the container.Votes: 0GitHub stars: 2
- Lm Studio Gpu InferenceCore skill for leveraging local GPU processing via LM Studio. Enables chat completions, model listing, and embedding generation through the OpenAI-compatible API at http://host.docker.internal:1234/v1. Use when projects need GPU-accelerated LLM inference or to run programs like OpenPlanter inside the container.Votes: 0GitHub stars: 2
- Real Time DataLive market prices for stocks, ETFs, indices via yfinance and cryptocurrency prices via ccxt. Use for current prices, historical OHLCV data, and market dashboards.Votes: 0GitHub stars: 2
- Self Optimizing SkillA skill that monitors its own execution patterns, logs token usage vs. success rate, and gradually optimizes its approach using RL-inspired feedback loopsVotes: 0GitHub stars: 2
- Self Optimizing SkillA skill that monitors its own execution patterns, logs token usage vs. success rate, and gradually optimizes its approach using RL-inspired feedback loopsVotes: 0GitHub stars: 2
- System Prompt EngineeringFramework for constructing effective system messages and prompts that define AI assistant behavior, capabilities, and response patterns. Use when configuring agent personality, tool usage, or behavioral guidelines.Votes: 0GitHub stars: 2
- Web Research MacroUnified web research pipeline combining browser navigation, content sanitization, and document extraction into a single call. Use when you need to extract clean text from websites requiring JavaScript rendering or login.Votes: 0GitHub stars: 2
- Web Research MacroWorkflow methodology for extracting clean text from websites. Sequences browser_agent, inline sanitization, and document_query into a repeatable pipeline. Not an importable module — a pattern for calling tools that already exist.Votes: 0GitHub stars: 2
- Api Caller'User asks to call an API, fetch data from a URL, make HTTP requests,Votes: 0GitHub stars: 2
- Api Rate LimitsThis skill contains knowledge about Anthropic's Claude API rate limitingVotes: 0GitHub stars: 2
- Artifact SavingWhen you produce a visual artifact (HTML dashboard, SVG diagram, interactiveVotes: 0GitHub stars: 2
- Code Execution Tool Import ErrorUse before calling code_execution_tool in a context that previously failed with 'import_error'. A required Python package is not installed in the active virtual environment. The correct pip binary is /opt/venv-a0/bin/pip \u2014 bare 'pip' may install into a difVotes: 0GitHub stars: 2
- Code Execution Tool Interactive PromptUse before calling code_execution_tool in a context that previously failed with 'interactive_prompt'. Command entered interactive mode requiring keyboard input. This execution environment cannot provide stdin input to running commands.Votes: 0GitHub stars: 2
- Code Execution Tool Terminal Session HungUse before calling code_execution_tool in a context that previously failed with 'terminal_session_hung'. A previous command is still running or hung in this terminal session. New commands cannot execute until the session is reset.Votes: 0GitHub stars: 2
- Search Engine Interactive PromptUse before calling search_engine in a context that previously failed with 'interactive_prompt'. Command entered interactive mode requiring keyboard input. This execution environment cannot provide stdin input to running commands.Votes: 0GitHub stars: 2
- Text Editor Interactive PromptUse before calling text_editor in a context that previously failed with 'interactive_prompt'. Command entered interactive mode requiring keyboard input. This execution environment cannot provide stdin input to running commands.Votes: 0GitHub stars: 2
- Text Editor Oversized Tool WriteUse before calling text_editor in a context that previously failed (oversized_tool_write). A text_editor:write was blocked because the content exceeds the ~5000-char JSON payload limit and would truncate, breaking the tool call.Votes: 0GitHub stars: 2
- Bdi Mental StatesImplement Belief-Desire-Intention (BDI) architecture for cognitive agents. Models rational agency through belief formation, goal management, and intention commitment cycles.Votes: 0GitHub stars: 2
- Claude Code Prompt Generation'Spec is complete and user is ready to hand off to Claude Code for implementation.Votes: 0GitHub stars: 2
- Command StructureUse this skill when designing, building, or operating multi-agent AIVotes: 0GitHub stars: 2
- Context CompressionCompress conversation history to reclaim token budget while preserving operational capability. Extracts durable facts to memory, discards noise, and injects structured summaries.Votes: 0GitHub stars: 2
- Context DegradationDiagnose and mitigate LLM context window degradation: lost-in-the-middle effects, instruction drift, capability regression, and tool-use failures. Provides diagnostic protocols, recovery strategies, prevention patterns, and Agent Zero integration procedures for maintaining agent performance across extended sessions.Votes: 0GitHub stars: 2
- Create SkillWizard for creating new Agent Zero skills. Guides users through creating well-structured SKILL.md files. Use when users want to create custom skills.Votes: 0GitHub stars: 2
- Cross Instance LearningTwo or more solutions to the same problem exist — either from parallelVotes: 0GitHub stars: 2
- Design Note WritingA new architectural concept has been identified but is not yet readyVotes: 0GitHub stars: 2
- Documentation SyncAfter any spec is created or updated, after new research is integrated,Votes: 0GitHub stars: 2
- Extension Debug DiagnosticsAn Agent-Zero extension is deployed but not producing expected behavior.Votes: 0GitHub stars: 2
- Human Captcha Intervention'Browser-based automation encounters a CAPTCHA challenge that cannotVotes: 0GitHub stars: 2
- Integration AssessmentUser discovers an external project (GitHub repo, tool, framework, service)Votes: 0GitHub stars: 2
- Investigation WorkflowThis skill orchestrates recursive OSINT investigations across heterogeneousVotes: 0GitHub stars: 2
- Investigative ReasoningUse this skill when investigating complex multi-source problems requiring entity resolution, cross-dataset linking, or public records correlation. Adapts OpenPlanter's methodology for Agent Zero investigations — joining disparate data sources through fuzzy matching and relationship inference. Triggers: 'investigate', 'find connections between', 'link these entities', 'cross-reference', 'entity resolution', 'follow the money'.Votes: 0GitHub stars: 2
- Irreversibility GateUse this skill before executing ANY action that interacts with externalVotes: 0GitHub stars: 2