diff --git a/README.md b/README.md
index dcaceb9..147938b 100644
--- a/README.md
+++ b/README.md
@@ -1,5 +1,7 @@
# miniCISO
+[**Visit the MiniCISO project site →**](https://icidade.github.io/miniCISO/)
+
Agentic security staff for [Hermes Agent](https://github.com/NousResearch/hermes-agent), distributed as a reproducible overlay of profiles, prompts, templates, and operating policies.
> This repository is **not a Hermes fork**. MiniCISO is a public overlay installed on top of a Hermes runtime pinned by version and commit in [`config/hermes-version.env`](config/hermes-version.env).
diff --git a/docs/.nojekyll b/docs/.nojekyll
new file mode 100644
index 0000000..8b13789
--- /dev/null
+++ b/docs/.nojekyll
@@ -0,0 +1 @@
+
diff --git a/docs/PUBLISHING.md b/docs/PUBLISHING.md
new file mode 100644
index 0000000..44446ca
--- /dev/null
+++ b/docs/PUBLISHING.md
@@ -0,0 +1,26 @@
+# Preview and publish the landing page
+
+The site is plain HTML, CSS, and JavaScript with no build step.
+
+## Preview locally
+
+From the repository root, run:
+
+```bash
+python -m http.server 8000 --directory docs
+```
+
+Then open `http://localhost:8000/`.
+
+## Publish with GitHub Pages
+
+After this change is merged:
+
+1. Open **Settings → Pages** in the repository.
+2. Under **Build and deployment**, choose **Deploy from a branch**.
+3. Select the `main` branch and the `/docs` folder.
+4. Save and verify `https://icidade.github.io/miniCISO/` after the deployment completes.
+
+GitHub Pages supports any source branch; `gh-pages` is a common convention, not a required branch. Keeping the source in `main/docs` makes content changes reviewable alongside the project documentation.
+
+All site assets use relative paths so the page works at the `/miniCISO/` project subpath. The canonical and social-preview URLs intentionally use the production URL.
diff --git a/docs/assets/diagram.png b/docs/assets/diagram.png
new file mode 100644
index 0000000..c5a5d47
Binary files /dev/null and b/docs/assets/diagram.png differ
diff --git a/docs/assets/favicon.jpg b/docs/assets/favicon.jpg
new file mode 100644
index 0000000..1f84330
Binary files /dev/null and b/docs/assets/favicon.jpg differ
diff --git a/docs/assets/hero-illustration.png b/docs/assets/hero-illustration.png
new file mode 100644
index 0000000..8112598
Binary files /dev/null and b/docs/assets/hero-illustration.png differ
diff --git a/docs/assets/logo-dark.png b/docs/assets/logo-dark.png
new file mode 100644
index 0000000..63cc699
Binary files /dev/null and b/docs/assets/logo-dark.png differ
diff --git a/docs/assets/social-preview.png b/docs/assets/social-preview.png
new file mode 100644
index 0000000..cb0acc7
Binary files /dev/null and b/docs/assets/social-preview.png differ
diff --git a/docs/index.html b/docs/index.html
new file mode 100644
index 0000000..465f61a
--- /dev/null
+++ b/docs/index.html
@@ -0,0 +1,128 @@
+
+
+
+
+
+ MiniCISO — Evidence-Driven Agentic Security Staff
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+ Skip to content
+
+
+
+
+
+
+
+
Open-source security orchestration
+
An evidence-driven security staff, orchestrated for real assessments.
+
MiniCISO coordinates specialized security agents for threat modeling, architecture, code review, AppSec, compliance, offensive validation, recon, and quality assurance—while keeping evidence, uncertainty, and human accountability visible.
Security work needs a chain of reasoning, not a pile of output.
Assessments break down when evidence is scattered, tools are disconnected, and early signals become premature conclusions. MiniCISO makes the operating discipline explicit.
+
+ 01
Evidence before narrative
Claims start from traceable artifacts. Scanner output and recon remain inputs—not findings by default.
+ 02
Specialists, not one generic prompt
The Chief of Staff delegates to focused SMEs whose scopes and boundaries are public.
+ 03
Explicit validation gates
Conclusions pass through adversarial checks and mandatory Security QA before delivery.
+ 04
Uncertainty stays visible
Assumptions, confidence, residual risk, and missing evidence remain part of the record.
+ 05
Human accountability
Authorization and consequential judgment stay with accountable security professionals.
+
+
+
+
+
02 / How it works
One engagement. Nine disciplined stages.
The Chief of Staff scopes the work, selects the right specialists, correlates their analysis, and enforces QA before synthesis.
+
+
01Intake & scope
02Evidence collection
03Hypothesis formation
04Selective retrieval
05SME analysis
06Cross-SME correlation
07Security QA
08Final synthesis
09Follow-up & lessons
+
+
+
+
+
+
03 / Meet the security staff
Coordinated roles. Clear boundaries.
Each profile has a public mission, defined scope, mandatory assumptions, confidence, and residual-risk reporting.
+
+
Coordinator
Chief of Staff
Scopes engagements, delegates work, correlates evidence, and owns the final synthesis.
Who should investigate? What is ready to deliver?
+
Specialist
Threat Modeling
Maps assets, trust boundaries, threats, and mitigations.
What can go wrong? Where are the critical paths?
+
Specialist
Security Architecture
Reviews system design, boundaries, controls, and resilience.
Do controls fit the design? Where does trust accumulate?
+
Specialist
Code Review
Inspects implementation evidence for security-relevant defects.
Is the weakness reachable? What code proves the claim?
+
Specialist
AppSec Assessment
Examines application risk across design, code, and deployment context.
What is exploitable? What evidence is still needed?
+
Specialist
Compliance Mapper
Maps technical evidence to relevant control requirements.
Which controls apply? What supports compliance?
+
Specialist
Offensive Security
Performs explicitly authorized validation within defined boundaries.
Can the path be validated safely? Is impact demonstrated?
+
Specialist
Recon & Attack Surface
Prioritizes exposed surfaces without turning discovery into findings.
What is exposed? What deserves controlled validation?
+
Quality gate
Security QA
Challenges evidence, impact, confidence, and reporting quality.
Is the conclusion supported? What would invalidate it?
+
Start with scope, evidence, and a clear objective.
A good first request names the target, boundaries, known constraints, available artifacts, and desired output. MiniCISO may pause to ask for what is missing.
The human operator stays outside—and above—the automated trust boundary.
MiniCISO is a public overlay installed on a pinned Hermes Agent runtime. It is not a Hermes fork. Profiles, policies, templates, and shared workspace coordination remain reproducible without publishing credentials or private runtime state.
The repository packages profiles, prompts, templates, operating policies, bootstrap and validation scripts, public documentation, and sanitized configuration.
Independent security engineering, built in public.
MiniCISO is an independent open-source security engineering project created by Irlan Cidade. It explores how specialized agents, explicit operating policies, evidence management, and human oversight can support more structured security assessments.
