ContextForge MCP Gateway

221 tools. 76 can modify or destroy data without limits.

2 destructive tools with no built-in limits. Policy required.

Last updated:

76 can modify or destroy data
145 read-only
221 tools total

Community server · catalogue entry verified 03/07/2026

How to control ContextForge MCP Gateway ↓

What ContextForge MCP Gateway exposes to your agents

Read (145) Write / Execute (74) Destructive / Financial (2)
Critical Risk

The most dangerous ContextForge MCP Gateway tools

76 of ContextForge MCP Gateway's 221 tools can modify, destroy, or commit something on every call — and an agent calls them with no built-in limits.

How to control ContextForge MCP Gateway

PolicyLayer is an MCP gateway — it sits between your AI agents and ContextForge MCP Gateway, and nothing reaches the server without passing your rules. These are the rules we recommend:

Deny destructive operations
{
  "clear_cache": {
    "deny_if": [
      {
        "conditions": [],
        "on_deny": "Blocked by default. Requires approval."
      }
    ]
  }
}

Destructive tools should never be available to autonomous agents without human approval.

Rate limit write operations
{
  "action_item_tracker": {
    "limits": [
      {
        "counter": "action_item_tracker_per_hour",
        "window": "hour",
        "max": 30,
        "scope": "grant"
      }
    ]
  }
}

Prevents bulk unintended modifications from agents caught in loops.

Cap read operations
{
  "agent.analyze_reasoning": {
    "limits": [
      {
        "counter": "agent.analyze_reasoning_per_minute",
        "window": "minute",
        "max": 60,
        "scope": "grant"
      }
    ]
  }
}

Controls API costs and prevents retry loops from exhausting upstream rate limits.

  1. Create a free account and register ContextForge MCP Gateway — nothing to install.
  2. Add these rules — paste them, or build them visually. Tune the limits to your setup.
  3. Point your MCP client (Claude, Cursor, anything) at your gateway URL.
ENFORCE POLICY ON CONTEXTFORGE MCP GATEWAY →

Instant setup, no code required.

All 221 ContextForge MCP Gateway tools

WRITE 56 tools
Write action_item_tracker Merge action item updates Write add_content Add content to a LaTeX document Write add_edge Add an edge to a DOT graph Write add_figure Add a figure to a LaTeX document Write add_formula Add a formula to a cell Write add_heading Add a heading to a document Write add_image Add an image to a slide Write add_node Add a node to a DOT graph Write add_numbers Add two numbers and return a structured result with output schema Write add_section Add a section to a LaTeX document Write add_shape Add a shape to a slide Write add_slide Add a new slide to the presentation Write add_table Add a table to a document Write add_text Add text to a document Write add_text_box Add a text box to a slide Write bitbucket_create_pr Create a new pull request Write bitbucket_merge_pr Merge a pull request Write bitbucket_pr_comments List or add comments on a pull request Write calibration.optimize_rubrics Tune evaluation rubrics for better alignment with human judgments Write communications_planner Plan communications cadence per stakeholder Write configure_model configure_model Write convert_batch Convert multiple documents to the same format Write convert_content Convert raw HTML or text content to markdown Write convert_document Convert a document to another format using LibreOffice Write create_bar_chart Create bar chart for categorical data Write create_chart Create a chart with flexible data input and configuration Write create_diagram Create and optionally render a Mermaid diagram Write create_document Create a new DOCX document Write create_flowchart Create flowchart from structured data Write create_from_template Create a LaTeX document from a template Write create_gantt_chart Create Gantt chart from task data Write create_graph Create a new DOT graph file Write create_line_chart Create line chart for time series or continuous data Write create_presentation Create a new PowerPoint presentation Write create_scatter_plot Create scatter plot with advanced customization Write create_sequence_diagram Create sequence diagram from participants and messages Write create_user Create a user profile with structured validation Write create_visualization Generate statistical plots and charts Write create_workbook Create a new XLSX workbook Write format_cells Format cells in a worksheet Write format_text Format text in a document Write generate_batch_qr_codes Generate multiple QR codes Write generate_dataset Generate a synthetic dataset and persist it for later retrieval Write generate_work_breakdown Generate a work breakdown structure from scope narrative. Write merge_documents Merge multiple documents into one file Write produce_gantt_diagram Generate gantt chart artefacts from schedule Write project_brief_generator Generate project brief summary Write risk_register_manager Manage and rank risks by severity Write save_presentation Save the presentation to a file Write set_attributes Set graph, node, or edge attributes Write set_slide_content Set the main content of a slide Write set_slide_title Set the title of a slide Write sprint_planning_helper Assemble sprint backlog based on capacity and priority Write stakeholder_matrix Produce stakeholder matrix diagram Write workflow.create_evaluation_suite Define comprehensive evaluation pipeline with multiple tools and success criteria Write write_data Write data to a worksheet
READ 145 tools
Read agent.analyze_reasoning Evaluate agent Read agent.evaluate_tool_use Assess agent Read agent.measure_task_completion Evaluate end-to-end task success against measurable criteria Read analyze_code Analyze code structure, complexity, and dependencies Read analyze_csv Perform automated analysis of CSV data Read analyze_dataset Perform comprehensive dataset analysis and profiling Read analyze_document Analyze document structure and content Read analyze_feed Comprehensive feed analysis with recommendations Read analyze_graph Analyze a DOT graph structure and metrics Read analyze_text Analyze text and recommend optimal chunking strategy Read analyze_workbook Analyze workbook content, structure, and formulas Read baseline_vs_actual Compare baseline vs actual metrics Read batch_convert Convert multiple URLs to markdown in parallel Read bias.assess_cultural_sensitivity Evaluate cross-cultural appropriateness and cultural awareness Read bias.detect_demographic_bias Identify bias against protected groups using pattern matching and LLM assessment Read bias.detect_linguistic_bias Identify language-based discrimination and dialect bias Read bias.evaluate_outcome_equity Check for disparate impacts across protected groups in outcomes Read bias.measure_intersectional_fairness Evaluate compound bias effects across multiple identity dimensions Read bias.measure_representation_fairness Assess balanced representation across groups in different contexts Read bitbucket_get_pipeline Get status and details for a specific pipeline run Read bitbucket_get_pr Get details for a specific pull request including diff Read bitbucket_get_repo Get details for a specific repository Read bitbucket_list_pipelines List pipeline runs for a repository Read bitbucket_list_prs List pull requests for a repository Read bitbucket_list_repos List repositories accessible to the authenticated user Read bitbucket_search_code Search code across repositories in a workspace Read calibration.test_judge_agreement Measure agreement between different judges and human evaluators Read change_request_tracker Summarise change request impacts Read chat_with_csv Chat with CSV data using natural language queries Read chunk_markdown Chunk markdown text with header awareness Read chunk_text Chunk text using various strategies (recursive, semantic, sentence, fixed_size) Read cluster_by_topic Cluster feed entries by semantic similarity Read convert_file Convert a local file to markdown format Read convert_url Convert URL content to markdown format with multiple engines and options Read decode_qr_code Decode QR code from image file Read divide_numbers Divide two numbers with error handling in output Read document_search document_search Read document_template_library Expose packaged PM templates Read earned_value_calculator Compute earned value management metrics Read echo Simple echo tool without output schema (for comparison) Read extract_classes Extract class definitions from code Read extract_functions Extract function definitions from code Read extract_text Extract all text content from a document Read fetch_rss Fetch and parse an RSS feed from URL Read filter_by_date Filter entries by date range Read find_by_author Find entries by author/speaker Read find_duplicates Find duplicate or near-duplicate entries using semantic similarity Read find_related_entries Find related entries to a specific entry Read fixed_size_chunk Fixed-size chunking with word boundary options Read get_capabilities Get information about converter capabilities and available engines Read get_csv_info Get comprehensive information about CSV data structure Read get_document_info Get information about a document Read get_feed_metadata Get feed metadata (title, description, author, etc.) Read get_feed_statistics Get comprehensive feed statistics and analysis Read get_latest_entries Get N most recent entries from feed Read get_model_info get_model_info Read get_presentation_info Get information about the presentation Read get_sandbox_info Get current sandbox capabilities and configuration Read get_server_info Get server information and capabilities Read get_strategies List available chunking strategies and capabilities Read get_supported_charts Get list of supported chart types and capabilities Read get_templates Get Mermaid diagram templates Read glossary_lookup Provide glossary definitions for common PM terms Read hybrid_search hybrid_search Read inspect_feed_schema Inspect feed schema and available fields Read judge.evaluate_response Evaluate a single response using LLM-as-a-judge with customizable criteria and rubrics Read judge.evaluate_with_reference Evaluate response against a gold standard reference using LLM-as-a-judge Read judge.pairwise_comparison Compare two responses and determine which is better using LLM-as-a-judge Read lessons_learned_catalog Aggregate lessons learned entries Read list_authors List all unique authors/speakers with entry counts Read list_categories List all categories/tags with entry counts Read list_generated_datasets List metadata about previously generated datasets Read list_layouts List available layout engines and output formats Read list_presets List available dataset presets with column definitions Read list_supported_formats List supported input and output formats Read load_dataset Load data from various sources and formats (CSV, JSON, Parquet, SQL, Excel) Read meeting_minutes_summarizer Summarise meeting transcript into decisions and actions Read multilingual.assess_cultural_adaptation Evaluate cultural appropriateness and localization quality Read multilingual.detect_language_mixing Identify inappropriate code-switching or language mixing Read multilingual.evaluate_translation_quality Assess translation accuracy and quality across languages Read multilingual.measure_cross_lingual_consistency Check consistency across multiple language versions Read multiply_numbers Multiply two numbers with structured output Read open_presentation Open an existing PowerPoint presentation Read performance.assess_computational_efficiency Measure resource usage and computational efficiency metrics Read performance.measure_response_latency Track generation speed and response times with statistical analysis Read performance.monitor_memory_usage Track memory consumption patterns during execution Read privacy.assess_consent_clarity Evaluate clarity and comprehensibility of consent notices Read privacy.assess_data_minimization Evaluate if data collection follows minimization principles Read privacy.assess_privacy_by_design Evaluate privacy-by-design implementation in systems Read privacy.detect_data_leakage Identify unintended data exposure or leakage in outputs Read privacy.detect_pii_exposure Detect personally identifiable information in text with configurable sensitivity Read privacy.evaluate_consent_compliance Assess consent mechanisms and compliance with privacy regulations Read privacy.evaluate_data_retention_compliance Assess data retention policy compliance and effectiveness Read privacy.measure_anonymization_effectiveness Evaluate effectiveness of data anonymization techniques Read project_health_dashboard Produce project health dashboard summary Read prompt.assess_relevance Measure semantic alignment between prompt and outputs using embeddings Read prompt.evaluate_clarity Assess prompt clarity using multiple rule-based and LLM-based metrics Read prompt.measure_completeness Evaluate if prompt generates complete responses covering expected components Read prompt.test_consistency Test prompt consistency across multiple runs and temperature settings Read quality.assess_toxicity Detect harmful or biased content using pattern matching and LLM analysis Read quality.evaluate_factuality Check factual accuracy of responses against knowledge bases Read quality.measure_coherence Analyze logical flow and consistency of text using rule-based and LLM metrics Read rag.assess_answer_groundedness Verify answers are grounded in provided context by checking claim support Read rag.assess_citation_accuracy Validate citation quality and accuracy against source documents Read rag.benchmark_retrieval_systems Compare different retrieval approaches using standard IR metrics Read rag.detect_hallucination_vs_context Identify when responses contradict provided context using statement verification Read rag.evaluate_retrieval_coverage Measure if key information was retrieved by checking topic coverage Read rag.evaluate_retrieval_relevance Assess relevance of retrieved documents to the query using semantic similarity and LLM judges Read rag.measure_chunk_relevance Evaluate individual chunk relevance scores using semantic similarity and LLM assessment Read rag.measure_context_utilization Check how well retrieved context is used in the generated answer Read read_data Read data from a worksheet Read resource_allocator Report resource allocation variance Read retrieve_dataset Retrieve persisted dataset content in CSV or JSONL format Read robustness.evaluate_prompt_injection_resistance Check prompt injection defenses and security measures Read robustness.measure_consistency_under_perturbation Check output stability under various input perturbations Read safety.assess_instruction_following Measure adherence to constraints and safety instructions Read safety.detect_harmful_content Identify dangerous or harmful outputs across multiple risk categories Read safety.evaluate_refusal_appropriateness Check when system appropriately refuses potentially harmful requests Read safety.measure_value_alignment Assess alignment with human values and ethical principles Read sample_data_catalog List packaged sample data assets Read schedule_optimizer Suggest lightweight schedule optimisations Read scope_guardrails Check proposed features against scope guardrails Read search_all Search across all fields (title, description, author, categories) Read search_descriptions Search RSS feed entries by description/content Read search_multi_field_semantic Multi-field semantic search (custom field selection) Read search_subtitles_semantic Semantic search in podcast subtitles (iTunes, Google Play, etc.) Read search_summaries_semantic Semantic search in episode summaries/descriptions Read search_titles Search RSS feed entries by title Read semantic_chunk Semantic chunking based on content similarity Read sentence_chunk Sentence-based chunking with configurable grouping Read server.get_available_judges Get list of available judge models and their capabilities Read server.get_cache_stats Get caching system statistics and performance metrics Read server.get_evaluation_results List evaluation results with optional filtering Read server.get_evaluation_suites List all created evaluation suites Read similarity_search Semantic similarity search using AI embeddings Read split_code Split code into logical segments using AST analysis Read status_report_generator Render status report markdown via template Read summarize_dataset Return summary statistics for a generated dataset Read time_series_analysis Analyze time series data patterns, trends, and forecasting Read validate_code Validate Python code without executing it Read validate_email Validate email address format Read validate_graph Validate DOT file syntax Read validate_mermaid Validate Mermaid diagram syntax Read validate_qr_data Validate and analyze QR code data before generation Read workflow.compare_evaluations Compare results across multiple evaluation runs with statistical analysis

Related servers

Other MCP servers with similar tools — same risk classification, starter policies for each.

Questions about ContextForge MCP Gateway

Can an AI agent delete data through the ContextForge MCP Gateway MCP server? +

Yes. The ContextForge MCP Gateway server exposes 2 destructive tools including clear_cache, delete_slide. These permanently remove resources with no undo. PolicyLayer blocks destructive tools by default so they never reach the upstream server.

How do I prevent bulk modifications through ContextForge MCP Gateway? +

The ContextForge MCP Gateway server has 56 write tools including action_item_tracker, add_content, add_edge. Set a rate limit in your policy -- for example, 10 calls per hour prevents an agent from making more than 10 modifications per hour. PolicyLayer enforces this at the gateway, before calls reach ContextForge MCP Gateway.

How many tools does the ContextForge MCP Gateway MCP server expose? +

221 tools across 4 categories: Destructive, Execute, Read, Write. 145 are read-only. 76 can modify, create, or delete data.

How do I enforce a policy on ContextForge MCP Gateway? +

Register the ContextForge MCP Gateway MCP server in PolicyLayer, apply the suggested rules above (adjust the limits to your use case), and point your AI client at the PolicyLayer proxy URL instead of the server directly. Your agents keep the same tools; PolicyLayer evaluates every call against policy before it executes. Nothing to install, live in minutes.

Enforce policy on every ContextForge MCP Gateway tool call.

Deterministic rules across all 221 ContextForge MCP Gateway tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.

Instant setup, no code required.

221 ContextForge MCP Gateway tools catalogued and risk-classified — across an index of 43,000+ MCP servers.

// GET IN TOUCH

Have a question or want to learn more? Send us a message.

Message sent.

We'll get back to you soon.