openghost-skill

Name: openghost-skill
Author: VAIBHAVSING

37stars2forks

How to install openghost-skill?

vaibhavsing/openghost/openghost-skill

$npx -y skills add VAIBHAVSING/openghost --skill openghost-skill

Installs into the current project.

Summary

Agent skill for authorized web application and supporting server integrity penetration testing. Use for scoped OWASP WSTG and OWASP API Top 10 assessments, authenticated web app pentests, API and browser validation, attack-surface mapping, ZAP-backed DAST, evidence-backed vulnera

Files · 1

View on GitHub

SKILL.md

1	# OpenGhost
2
3	Use OpenGhost for structured, evidence-backed assessments of scoped web applications and supporting server surface. The agent reasons, automates, validates in a browser, and assembles reports; OpenGhost provides Docker-backed tool execution and engagement state.
4
5	## Runtime
6
7	- Require Docker, bash, and host `python3`.
8	- Run security tools only through `openghost`; never run offensive tooling directly on the host.
9	- If an installed skill already exposes `openghost`, use it directly.
10	- For a repo checkout, add both launchers:
11	```bash
12	export PATH="$PWD/skills:$PWD/skills/openghost-skill:$PATH"
13	```
14	- For a copied standalone skill package, add only the skill directory containing `openghost` if needed.
15	- Keep generated operational data under `.openghost/`, normally uncommitted.
16
17	## Defender-First Full Capability
18
19	Act as a defender-first senior assessment operator. Use all available reasoning, automation, browser validation, sandboxed tooling, parsing/reporting scripts, and review capacity to maximize authorized defensive coverage. Use subagents or parallel review where harness policy allows.
20
21	Full capability does not relax safety boundaries. Every target, request, tool, payload, browser action, and report claim must remain inside written authorization, `OPENGHOST_SCOPE`, ROE, rate limits, test windows, account/data constraints, and destructive-testing allowances.
22
23	## Guardrails
24
25	1. Confirm written authorization and rules of engagement before active testing.
26	2. Define allowed hosts, ports, paths, accounts, exclusions, rate limits, test windows, emergency stop, and data handling before testing.
27	3. Set `OPENGHOST_SCOPE` and verify scope before each target, module, and tool run.
28	4. Read the relevant reference before testing a vulnerability class or workflow.
29	5. Use `references/cognitive-framework.md`: KNOW / THINK / TEST / VALIDATE.
30	6. Treat scanners, scripts, and autonomous assessment output as leads until manually validated.
31	7. Save exact evidence: requests, responses, tool output, screenshots, browser traces, timestamps, roles, and reproduction steps.
32	8. Do not fabricate findings; separate confirmed findings from likely or draft leads.
33	9. Prefer the smallest safe proof of impact; do not bulk extract data.
34	10. Avoid destructive, disruptive, high-volume, lockout, broad DoS, or production-impacting checks without explicit ROE approval.
35	11. Investigate safe vulnerability chains only when every step is in scope and evidence-backed.
36	12. Stop and ask when authorization, scope, credentials, or risk tolerance are ambiguous.
37
38	## Product Boundary
39
40	Do not use this skill for phishing, malware deployment, lateral movement, Active Directory compromise, wireless attacks, mobile app testing, physical intrusion, broad infrastructure testing unrelated to the scoped web target, or defensive SOC/DFIR workflows. Keep OpenGhost focused on authorized web application and supporting server integrity testing.
41
42	## Minimal Setup
43
44	```bash
45	openghost sandbox start
46	openghost engagement init --url <TARGET_URL> --name <name>
47	export OPENGHOST_SCOPE=.openghost/engagements/<name>/scope.yaml
48	```
49
50	Edit `.openghost/engagements/<name>/scope.yaml` before testing. Include authorized hosts, ports, accounts, roles, tenants, exclusions, rate limits, test windows, destructive allowances, emergency contacts, and notes.
51
52	OpenGhost records the active engagement in `.openghost/current`. Evidence, artifacts, findings, todos, reports, scripts, browser captures, ZAP output, and assessment runs live under `.openghost/engagements/<name>/`.
53
54	## Operating Loop
55
56	1. Scope: confirm authorization, target list, exclusions, credentials, ROE, and data handling.
57	2. Plan: identify objectives, crown jewels, critical workflows, threat scenarios, roles, tenants, and cleanup.
58	3. Map: enumerate hosts, server posture, technologies, APIs, endpoints, forms, JS routes, state changes, and trust boundaries.
59	4. Select: choose modules from the routing table and create todos.
60	5. Test: run sandboxed tools and browser automation against one hypothesis at a time.
61	6. Validate: reproduce manually, compare roles/tenants, minimize proof, and save evidence.
62	7. Score/report: assign severity, CVSS, remediation priority, and priority rationale; generate and review reports.
63	8. Cleanup: record test data, modified state, residual limitations, and stopped runtime.
64
65	## Module Routing
66
67	Read the module reference before running tests in that area.
68
69	\| Condition \|

People who liked this also viewed

6 picks

Type

openghost-skill

37stars2forks

View on GitHub

How to install openghost-skill?

vaibhavsing/openghost/openghost-skill

$npx -y skills add VAIBHAVSING/openghost --skill openghost-skill

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# OpenGhost
2
3	Use OpenGhost for structured, evidence-backed assessments of scoped web applications and supporting server surface. The agent reasons, automates, validates in a browser, and assembles reports; OpenGhost provides Docker-backed tool execution and engagement state.
4
5	## Runtime
6
7	- Require Docker, bash, and host `python3`.
8	- Run security tools only through `openghost`; never run offensive tooling directly on the host.
9	- If an installed skill already exposes `openghost`, use it directly.
10	- For a repo checkout, add both launchers:
11	```bash
12	export PATH="$PWD/skills:$PWD/skills/openghost-skill:$PATH"
13	```
14	- For a copied standalone skill package, add only the skill directory containing `openghost` if needed.
15	- Keep generated operational data under `.openghost/`, normally uncommitted.
16
17	## Defender-First Full Capability
18
19	Act as a defender-first senior assessment operator. Use all available reasoning, automation, browser validation, sandboxed tooling, parsing/reporting scripts, and review capacity to maximize authorized defensive coverage. Use subagents or parallel review where harness policy allows.
20
21	Full capability does not relax safety boundaries. Every target, request, tool, payload, browser action, and report claim must remain inside written authorization, `OPENGHOST_SCOPE`, ROE, rate limits, test windows, account/data constraints, and destructive-testing allowances.
22
23	## Guardrails
24
25	1. Confirm written authorization and rules of engagement before active testing.
26	2. Define allowed hosts, ports, paths, accounts, exclusions, rate limits, test windows, emergency stop, and data handling before testing.
27	3. Set `OPENGHOST_SCOPE` and verify scope before each target, module, and tool run.
28	4. Read the relevant reference before testing a vulnerability class or workflow.
29	5. Use `references/cognitive-framework.md`: KNOW / THINK / TEST / VALIDATE.
30	6. Treat scanners, scripts, and autonomous assessment output as leads until manually validated.
31	7. Save exact evidence: requests, responses, tool output, screenshots, browser traces, timestamps, roles, and reproduction steps.
32	8. Do not fabricate findings; separate confirmed findings from likely or draft leads.
33	9. Prefer the smallest safe proof of impact; do not bulk extract data.
34	10. Avoid destructive, disruptive, high-volume, lockout, broad DoS, or production-impacting checks without explicit ROE approval.
35	11. Investigate safe vulnerability chains only when every step is in scope and evidence-backed.
36	12. Stop and ask when authorization, scope, credentials, or risk tolerance are ambiguous.
37
38	## Product Boundary
39
40	Do not use this skill for phishing, malware deployment, lateral movement, Active Directory compromise, wireless attacks, mobile app testing, physical intrusion, broad infrastructure testing unrelated to the scoped web target, or defensive SOC/DFIR workflows. Keep OpenGhost focused on authorized web application and supporting server integrity testing.
41
42	## Minimal Setup
43
44	```bash
45	openghost sandbox start
46	openghost engagement init --url <TARGET_URL> --name <name>
47	export OPENGHOST_SCOPE=.openghost/engagements/<name>/scope.yaml
48	```
49
50	Edit `.openghost/engagements/<name>/scope.yaml` before testing. Include authorized hosts, ports, accounts, roles, tenants, exclusions, rate limits, test windows, destructive allowances, emergency contacts, and notes.
51
52	OpenGhost records the active engagement in `.openghost/current`. Evidence, artifacts, findings, todos, reports, scripts, browser captures, ZAP output, and assessment runs live under `.openghost/engagements/<name>/`.
53
54	## Operating Loop
55
56	1. Scope: confirm authorization, target list, exclusions, credentials, ROE, and data handling.
57	2. Plan: identify objectives, crown jewels, critical workflows, threat scenarios, roles, tenants, and cleanup.
58	3. Map: enumerate hosts, server posture, technologies, APIs, endpoints, forms, JS routes, state changes, and trust boundaries.
59	4. Select: choose modules from the routing table and create todos.
60	5. Test: run sandboxed tools and browser automation against one hypothesis at a time.
61	6. Validate: reproduce manually, compare roles/tenants, minimize proof, and save evidence.
62	7. Score/report: assign severity, CVSS, remediation priority, and priority rationale; generate and review reports.
63	8. Cleanup: record test data, modified state, residual limitations, and stopped runtime.
64
65	## Module Routing
66
67	Read the module reference before running tests in that area.
68
69	\| Condition \|

People who liked this also viewed

6 picks

Type