# io.github.runsec-io/mcp (npm · @runsec/mcp)

RunSec MCP server for workspace security scanning and remediation workflows.

- Trust score: 60/100 (medium)
- Change this week: +21
- Registry status: active
- Liveness: live
- Owner verified: no
- Last scored: 2026-08-03

## Components

- npm · `@runsec/mcp`: 60/100 (this document), [markdown](https://verifymcp.io/servers/runsec-io-mcp/runsec-mcp.md), [page](https://verifymcp.io/servers/runsec-io-mcp/runsec-mcp)

## Channel facts

- Registry: `npm`
- Package: `@runsec/mcp`
- Version: `1.0.127`
- Transport: `stdio`

## Trust breakdown

How this component scores in each security and reliability category. Every signal is checked automatically from public evidence about the published package, including repeated runs of it in an isolated sandbox, and we only credit what we can confirm. Scores are 0–100 per category. Scoring method: https://verifymcp.io/docs/scoring (what has changed: https://verifymcp.io/docs/scoring/changelog)

Scored 2026-08-03.

- **Supply Chain Security**: 88/100
  - No malware found by supply-chain analysis.
  - Only part of the dependency tree could be resolved (109 of 117), so this covers what we could see, not the whole tree.
  - No install/post-install scripts declared.
  - Only part of the dependency tree could be resolved (109 of 117), so this covers what we could see, not the whole tree.
- **Provenance & Transparency**: 19/100
  - Repository check failed: no source repository is declared.
  - Provenance check failed: no build-provenance attestation is published.
  - Clear OSI-approved license (MIT).
  - Actively maintained (last published 66 days ago).
  - Security-disclosure policy not yet verified: we couldn't inspect the source repository.
- **Schema Quality & AI Usability**: 56/100
  - AI-judged instruction clarity (fair).
  - Context-footprint check failed: tool/resource definitions use about 2352 tokens (~123/item across 19 items; 19 tools + 0 resources), over budget; trim descriptions and params.
  - Usage-examples check failed: none of the tools include examples.
- **Stability & Change Management**: 23/100
  - Stability observed for 7 of 30 days with no destabilising changes; credit accrues until the full window elapses.
- **Tool Coverage**: 100/100
  - 100% of tools have a non-trivial description (not blank, and not just the tool's name).
  - 100% of tool parameters carry a description.
- **Capabilities**: 100/100
  - Implements a supported MCP spec version (2025-11-25); the latest is 2026-07-28.

## Install

### Claude

```bash
claude mcp add runsec-io-mcp -- npx -y @runsec/mcp
```

### Codex

```bash
codex mcp add runsec-io-mcp -- npx -y @runsec/mcp
```

### opencode

```json
{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "runsec-io-mcp": {
      "type": "local",
      "command": [
        "npx",
        "-y",
        "@runsec/mcp"
      ],
      "enabled": true
    }
  }
}
```

### OpenClaw

```bash
openclaw mcp add runsec-io-mcp --command npx --arg -y --arg @runsec/mcp
```

### Hermes

```yaml
mcp_servers:
  runsec-io-mcp:
    command: "npx"
    args: ["-y", "@runsec/mcp"]
```

### Other

```json
{
  "mcpServers": {
    "runsec-io-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "@runsec/mcp"
      ]
    }
  }
}
```

## Changelog

Every change recorded for this component, newest first. Days that predate change tracking, or that we cannot explain, say so: "we were watching and nothing happened" and "we were not watching" are different claims.

### 2026-08-03 (score 60, +4)

- [functional improvement] Stability: unverified → 0.23

### 2026-08-02 (score 56, +30)

- [security regression] Provenance: unverified → fail
- [security improvement] Known CVEs: unverified → partial
- [security improvement] Install scripts: unverified → pass
- [security improvement] Malware scan: unverified → pass
- [security] Stability: Stability not yet verified: we do not have a sandbox capture of the MCP schema this version of the package serves yet.
- [functional regression] Tool coverage: 100 → unverified
- [functional regression] Capabilities: pass → unverified
- [functional improvement] Maintenance: unverified → pass
- [functional improvement] License: unverified → pass
- [functional improvement] Dependency health: unverified → partial
- [functional] Licence: MIT

### 2026-08-01 (score 26, +26)

- [security] Stability: Stability not yet verified: not enough scan history yet (needs a 30-day window).
- [functional improvement] Tool coverage: unverified → 100
- [functional improvement] MCP protocol: unverified → pass

### 2026-07-31 (score 0, −68)

- [functional] We updated how we score, so this day's move reflects our rubric, not a change to the server

### 2026-07-30 (score 68, +8)

- [functional improvement] Schema quality: unverified → fair

### 2026-07-29 (score 60, +42)

- [security regression] Provenance: unverified → fail
- [security improvement] Known CVEs: unverified → partial
- [security improvement] Install scripts: unverified → pass
- [functional improvement] Maintenance: unverified → pass
- [functional improvement] License: unverified → pass
- [functional improvement] Tool coverage: unverified → 100
- [functional] Licence: MIT

### 2026-07-28 (score 18, −21)

- [functional regression] Tool coverage: 100 → unverified
- [functional improvement] Dependency health: unverified → partial
- [functional] First check of Schema quality: unverified

### 2026-07-27 (score 39)

First indexed and scored.

## MCP tools (19)

### `runsec_audit_owasp` (~62 tokens)

Run OWASP audit against workspace files. Writes results to runsec-report.md.

Input parameters:

- `target_files` (array): Optional relative file paths to scan instead of full workspace.
- `workspace_path` (string, required): Absolute or relative path to repository root.

### `runsec_audit_pcidss` (~72 tokens)

Run PCI-DSS v4.0 Req 6.5 audit against workspace files. Writes results to runsec-report.md.

Input parameters:

- `target_files` (array): Optional relative file paths to scan instead of full workspace.
- `workspace_path` (string, required): Absolute or relative path to repository root.

### `runsec_audit_soc2` (~71 tokens)

Run SOC2 logical-access audit (JWT/session + RBAC patterns) against workspace files. Writes results to runsec-report.md.

Input parameters:

- `target_files` (array): Optional relative file paths to scan instead of full workspace.
- `workspace_path` (string, required): Absolute or relative path to repository root.

### `runsec_audit_hipaa` (~73 tokens)

Run HIPAA safeguards audit (PHI/PII logging + integrity) against workspace files. Writes results to runsec-report.md.

Input parameters:

- `target_files` (array): Optional relative file paths to scan instead of full workspace.
- `workspace_path` (string, required): Absolute or relative path to repository root.

### `runsec_audit_general` (~122 tokens)

Full workspace security audit (user intent: @runsec audit). Runs static engines plus mandatory agent architectural review (missing controls: auth, RBAC, rate limiting, CSRF, secrets-by-design). Writes runsec-semantic-brief.json; execute agent_instructions (7 phases incl. Step 2.5) and submit_agent_findings. Hub sync when API key is set.

Input parameters:

- `target_files` (array): Optional relative file paths to scan instead of full workspace.
- `workspace_path` (string, required): Absolute or relative path to repository root.

### `runsec_ignore_finding` (~175 tokens)

Persist an accepted false positive to .runsecignore.yaml (rule_id + file_path + line SHA256). Matching findings are excluded from future Semgrep audits in this workspace.

Input parameters:

- `file_path` (string, required): Relative path to the file within the workspace.
- `line_content` (string): Exact source line at the finding (used to compute SHA256 fingerprint).
- `line_sha256` (string): Precomputed SHA256 of line_content; use instead of line_content when already known.
- `reason` (string): Optional justification stored in .runsecignore.yaml.
- `rule_id` (string, required): Semgrep rule id or metric token (e.g. PY-001, runsec.domain-input-validation.cwe-89).
- `workspace_path` (string, required): Repository root where .runsecignore.yaml will be written.

### `runsec_list_skills` (~34 tokens)

List RunSec security skills (language/framework packs) with activation triggers, extensions, and pattern counts for remediation routing.

### `runsec_get_context` (~108 tokens)

Return index.md and patterns.md for a skill_id, plus optional RAG hints and skill orchestration for a file or question.

Input parameters:

- `file_content` (string): Optional file snippet for orchestration scoring.
- `file_path` (string): Optional file path for skill orchestration scoring.
- `question` (string): Optional question to fetch RAG hints within the skill.
- `skill_id` (string, required): Skill identifier (e.g. fastapi-async, nodejs-nestjs).

### `runsec_ask_guidance` (~61 tokens)

Semantic search over RunSec safe-pattern knowledge base. Returns top safe patterns and fix templates for a vulnerability description (anti-hallucination remediation).

Input parameters:

- `question` (string, required): Vulnerability description, metric id, or remediation question (natural language).

### `runsec_threat_model` (~115 tokens)

STRIDE architectural threat model: analyzes Syft SBOM, package.json, docker-compose, and repo signals. Writes runsec-threat-model.md and .threat-model-cache.json in the workspace.

Input parameters:

- `context` (string): Optional free-text context (stack, deployment model, integrations) to refine STRIDE analysis.
- `project_name` (string): Optional project name used to infer architecture profile (kubernetes-api, desktop-gateway, generic).
- `workspace_path` (string, required): Absolute or relative path to repository root.

### `runsec_update_threat_model` (~125 tokens)

Save IDE-agent STRIDE threat model to <workspace>/.runsec-cache/threat-model.json for reuse on the next audit (delta-oriented Step 2 in agent contract).

Input parameters:

- `profile` (string): Optional architecture profile label (e.g. kubernetes-api, generic).
- `scope_files` (array): Optional relative paths covered by this threat model revision.
- `threat_model_markdown` (string, required): Full or delta-merged STRIDE threat model markdown produced by the IDE agent.
- `workspace_path` (string, required): Absolute or relative path to repository root.

### `runsec_generate_review` (~133 tokens)

Penetration-tester / auditor report: runs a full GENERAL unified scan plus STRIDE threat modeling, then writes runsec-security-review.md (executive summary, business-risk findings, audit exclusions). English-only; distinct from runsec-report.md.

Input parameters:

- `context` (string): Optional deployment context (stack, integrations, regulatory scope) for STRIDE profiling.
- `project_name` (string): Optional display name for the security review document header.
- `target_files` (array): Optional relative file paths to scan instead of the full workspace.
- `workspace_path` (string, required): Absolute or relative path to repository root.

### `runsec_get_semantic_targets` (~133 tokens)

Return high-value authorization/business-logic review targets, expert prompts (DRF/FastAPI/Flutter baselines), and JSON finding schema for the IDE agent. No remote LLM calls — the Cursor agent performs semantic review locally.

Input parameters:

- `context` (string): Optional deployment/architecture notes to refine profile detection.
- `include_static_baseline` (boolean): When true (default), run a quick static merge pass to rank targets against Semgrep hits.
- `target_files` (array): Optional relative file paths to scope target selection.
- `workspace_path` (string, required): Absolute or relative path to repository root.

### `runsec_apply_remediation` (~162 tokens)

Apply a deterministic safe-pattern patch to a file. Verifies target_line_or_block matches exactly, validates replacement against Skills catalog, writes a .bak backup, then patches the file.

Input parameters:

- `file_path` (string, required): Relative path to the file within the workspace.
- `replacement_code` (string, required): Approved secure replacement; must align with safe_pattern/fix_template from runsec_ask_guidance.
- `rule_id` (string, required): Metric or rule id from RunSec skills (e.g. DVS-001, PY-001, DOCK-010).
- `target_line_or_block` (string, required): Exact line or multi-line block currently in the file (must match before patch).
- `workspace_path` (string, required): Repository root containing the target file.

### `execute_ephemeral_script` (~224 tokens)

Run a polyglot ephemeral security test: writes content to <workspace>/.runsec-temp/, executes command from workspace root (10s timeout), returns stdout/stderr, then deletes the script. Use {file} in command for script path.

Input parameters:

- `command` (string, required): Shell command to run from workspace root. Use {file} or {filepath} for the script path.
- `content` (string, required): Full source code of the ephemeral test (mock DB / in-memory only).
- `filename` (string, required): Basename for the ephemeral test file (e.g. test_bola_poc.py, poc_test.go).
- `finding_id` (string, required): The ID of the finding you are trying to verify (e.g. django-cors-wildcard). This MUST match the id you will use in submit_agent_findings.
- `timeout_ms` (number): Optional timeout in milliseconds (default 10000).
- `workspace_path` (string, required): Repository root; script is written to .runsec-temp/ under this path and command runs with cwd here.

### `execute_live_poc` (~229 tokens)

Execute a live HTTP request for API/network PoC validation (10s default timeout). Returns status, headers, and body (truncated at 256KB). Include finding_id matching submit_agent_findings — any completed HTTP response (including 4xx/5xx) records the PoC as executed for VERIFIED.

Input parameters:

- `body` (string): Optional request body (raw string; use with POST/PUT/PATCH).
- `finding_id` (string, required): The ID of the finding you are verifying (e.g. api-bola-idor). Must match the id used in submit_agent_findings.
- `headers` (object): Optional request headers as string key/value pairs.
- `method` (string, required): HTTP method: GET, POST, PUT, PATCH, DELETE, HEAD, or OPTIONS.
- `timeout_ms` (number): Optional timeout in milliseconds (default 10000, max 60000).
- `url` (string, required): Full http:// or https:// URL to request (including path and query).
- `workspace_path` (string, required): Repository root (for consistency with other verification tools).

### `report_cognitive_feedback` (~149 tokens)

Train the adaptive cognitive filter: record a False Positive (FP → AI_VETOED) or True Positive (TP → CONFIRMED_TP) for a static hint fingerprint_hash from runsec-semantic-brief.json. After two FP vetoes the finding is auto-suppressed on future scans.

Input parameters:

- `fingerprint_hash` (string, required): SHA-256 fingerprint from static_hints[].fingerprint_hash in the semantic brief.
- `reason` (string): Optional rationale stored with the feedback entry.
- `verdict` (string, required): FP = agent vetoed false positive; TP = confirmed true positive.
- `workspace_path` (string, required): Repository root where .runsec-adaptive-filter.json is stored.

### `runsec_get_active_findings` (~73 tokens)

Retrieve active/open vulnerabilities from sealed .runsec-cache/report.json (user intent: @runsec fix active findings). Returns unsealed agent_findings and static_snapshot summary — use instead of reading report.json manually.

Input parameters:

- `workspace_path` (string, required): Repository root where .runsec-cache/report.json is stored.

### `submit_agent_findings` (~231 tokens)

Canonical findings store + Hub sync (user intent: @runsec verify and sync). Default: merge by id (partial submits update only listed findings; others preserved). Use replace_all:true for full audit terminal submit. Remove fixed items via resolved_ids or finding resolution: resolved|false_positive. verification_status enum is strict; poc_output is mandatory on upserts.

Input parameters:

- `findings` (array, required): Structured findings only — server renders Markdown. Do not submit pre-formatted report text.
- `is_new_audit` (boolean): Set true on the first submit after runsec_audit_* to force replace_all (clears ghost findings from prior audits).
- `replace_all` (boolean): When true, incoming findings replace the entire cache (minus resolved_ids). Default false: merge by id. Auto-enabled when static audit is newer than last agent sync.
- `resolved_ids` (array): Finding ids to remove from cache (fixed/false positive). Use with partial verify submits; omit ids you want to keep unchanged.
- `workspace_path` (string, required): Repository root where runsec-report.md will be updated.

## Diagnostics

Captured diagnostic sections: Provenance, Dependencies. The full working is on the page: https://verifymcp.io/servers/runsec-io-mcp/runsec-mcp#diagnostics

## Score history

- 2026-08-03: 60
- 2026-08-02: 56
- 2026-08-01: 26
- 2026-07-31: 0
- 2026-07-30: 68
- 2026-07-29: 60
- 2026-07-28: 18
- 2026-07-27: 39

## Links

- npm package: https://www.npmjs.com/package/@runsec/mcp
- Socket report: https://socket.dev/npm/package/@runsec/mcp
- Changelog RSS feed: https://verifymcp.io/servers/runsec-io-mcp/runsec-mcp/changelog.xml
- Changelog JSON feed: https://verifymcp.io/servers/runsec-io-mcp/runsec-mcp/changelog.json
- HTML version of this page: https://verifymcp.io/servers/runsec-io-mcp/runsec-mcp