# AI use cases for Blitzstrike (shinthink/blitzstrike) Repo: [shinthink/blitzstrike](https://github.com/shinthink/blitzstrike) | Use cases: 4 | Page: https://dedupedai.com/use-cases/shinthink-blitzstrike/ Blitz Strike is a TypeScript MCP server that packages a three-tier security audit (recon, source analysis, live validation) as tools an AI agent can call. This file holds every use case of this repo. Each one starts with a "## " heading and ends with its URL. - Triage a source tree for reachable, unauthenticated sinks: https://dedupedai.com/use-cases/shinthink-blitzstrike/triage-source-tree-for-reachable-sinks/ - Verify live web findings with marker and control tests: https://dedupedai.com/use-cases/shinthink-blitzstrike/verify-live-web-findings-before-reporting/ - Score, dedupe, and report findings for a write-up: https://dedupedai.com/use-cases/shinthink-blitzstrike/score-dedupe-and-report-findings/ - Look up CVEs and known issues for a tech stack: https://dedupedai.com/use-cases/shinthink-blitzstrike/look-up-cves-for-a-stack/ ## How to triage a source tree for reachable, unauthenticated sinks with Blitzstrike Repo: [shinthink/blitzstrike](https://github.com/shinthink/blitzstrike) | Difficulty: Intermediate | Time: 15 minutes | Basis: From the README Let an agent scan your code and report sinks that are reachable without auth, with file and line references. **Who it is for:** A developer or reviewer preparing a codebase for release who needs a shorter list than a plain pattern scanner gives. **The problem:** Pattern matching alone flags many lines that are never reachable from an outside request, which wastes review time. **The outcome:** A list of findings with file:line references, traced from source to sink and enriched with escalation chains where a match exists. ### Steps 1. Run `npx -y blitzstrike doctor` to verify the runtime and the tool catalog. 2. Run `npx -y blitzstrike install` to register Blitz Strike with the agent CLIs it detects. 3. Restart your agent and ask it to "audit ./src" or point it at your source tree path. 4. Read each returned finding and open the referenced file and line to confirm the sink is reachable. 5. Keep only the findings you can explain, and note which ones still need a live test. ### Prompt to paste into your agent ```text Using the blitzstrike MCP server, call run_engagement with target and max_files . Return the findings grouped by chain, each with file and line references. Verify that every finding names a source and a sink that the tool traced, not only a matched pattern. ``` ### Caveat Blitz Strike findings are hypotheses from static analysis and should be confirmed by reading the code and, where relevant, a live test. Dual-use notice: only use this on software, sites and systems you own or are authorized to test. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/shinthink-blitzstrike/triage-source-tree-for-reachable-sinks/ ## How to verify live web findings with marker and control tests with Blitzstrike Repo: [shinthink/blitzstrike](https://github.com/shinthink/blitzstrike) | Difficulty: Advanced | Time: 30 minutes | Basis: From the README Check a scope, then let the agent confirm live findings with marker reflection and a negative control. **Who it is for:** A penetration tester with written authorization to test a specific web application. **The problem:** A scan hit on a live site may be a false positive, and an unverified finding should not reach a client report. **The outcome:** Each live finding carries a verdict from a live test, and out-of-scope targets are blocked before any request is sent. ### Steps 1. Run `npx -y blitzstrike doctor` to confirm the environment is ready. 2. Ask the agent to call scope_check with your target and written scope before any active testing. 3. Ask the agent to audit the authorized URL, for example "audit https://". 4. Ask the agent to run strike_verify on each candidate, with a marker, a negative control, and a baseline. 5. Use strike_resolve to record the verdict on each finding before it is written up. ### Prompt to paste into your agent ```text Using the blitzstrike MCP server, call scope_check for target with scope , then run_engagement on that target. For each candidate, call strike_verify and report only the findings with a live verdict. Verify that no request was sent to a host outside the scope. ``` ### Caveat The README describes live verification and scope enforcement, but the exact scope format and mode values are not shown in the README excerpt, so check the docs before a real engagement. Risk, reverse engineering: Active live verification sends requests to a real system, so run it only against targets you are authorized to test and within the agreed scope. Dual-use notice: only use this on software, sites and systems you own or are authorized to test. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/shinthink-blitzstrike/verify-live-web-findings-before-reporting/ ## How to score, dedupe, and report findings for a write-up with Blitzstrike Repo: [shinthink/blitzstrike](https://github.com/shinthink/blitzstrike) | Difficulty: Intermediate | Time: 15 minutes | Basis: From the README Turn a raw findings list into a deduplicated, CVSS-scored report with an integrity hash. **Who it is for:** A security engineer who has a list of findings from a scan and needs a clean report for a ticket or client. **The problem:** Several findings often share one root cause, and severity scores are slow to compute by hand. **The outcome:** A markdown or JSON report with one group per root cause, deterministic CVSS scores, and a SHA-256 integrity hash. ### Steps 1. Run `npx -y blitzstrike install` if the MCP server is not yet registered with your agent. 2. Give the agent your findings list in the format the tools accept. 3. Ask the agent to call dedup_findings so findings sharing a sink, source, and CWE collapse into one group. 4. Ask the agent to call cvss_score for each root cause using its attack vector values. 5. Ask the agent to call generate_report with format set to markdown or json. ### Prompt to paste into your agent ```text Using the blitzstrike MCP server, call dedup_findings on , then cvss_score for each group, then generate_report in . Return the report with a summary and the SHA-256 integrity hash. Verify that every group has a CVSS base score and severity. ``` ### Caveat The CVSS score is computed by the tool from the inputs you give it, so the input values still need to be judged by a person. Dual-use notice: only use this on software, sites and systems you own or are authorized to test. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/shinthink-blitzstrike/score-dedupe-and-report-findings/ ## How to look up CVEs and known issues for a tech stack with Blitzstrike Repo: [shinthink/blitzstrike](https://github.com/shinthink/blitzstrike) | Difficulty: Intermediate | Time: 15 minutes | Basis: From the README Query NVD and the tech correlation data to see which known issues match a product or version you run. **Who it is for:** A developer or operator who needs to know whether a technology in use has known CVEs before patching. **The problem:** Checking each product and CVE by hand across several sources takes time and is easy to skip. **The outcome:** A list of CVE details from NVD and the known vulnerabilities linked to a technology, to guide patch priority. ### Steps 1. Run `npx -y blitzstrike install` to register the MCP server with your agent. 2. Ask the agent to call nvd_lookup with a CVE ID you want to check, which needs no API key. 3. Ask the agent to call tech_correlation with the technology name from your stack. 4. Ask the agent to call cve_correlation for any CVE that appears in the results. 5. Record the CVEs that match your versions and decide which to patch first. ### Prompt to paste into your agent ```text Using the blitzstrike MCP server, call nvd_lookup for and tech_correlation for . Return the matching CVEs with their severity and the related product. Verify that each listed CVE is confirmed against the version you run before you act on it. ``` ### Caveat Correlation data is a starting point, so confirm each CVE against the exact product version you run in your environment. Dual-use notice: only use this on software, sites and systems you own or are authorized to test. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/shinthink-blitzstrike/look-up-cves-for-a-stack/