AGENT SKILL

samm-maturity

OWASP SAMM maturity self-assessment producing a scored 15-stream posture roadmap across five business functions with three-level maturity rubric.

License 许可
AGPL-3.0
最近更新
2026-11-14
Stars
★ 25
Forks
9

01 前言 Front Matter

name
samm-maturity
owner
pm
scope
co-security
status
active
version
1.1.0
metadata
{ "type": "process", "triggers": [ "samm", "samm assessment", "maturity assessment", "security posture assessment" ] }
description
OWASP SAMM maturity self-assessment producing a scored 15-stream posture roadmap across five business functions with three-level maturity rubric.
last_reviewed
1793491200000
prerequisites
Repository access, security artifacts (policies, scan results, incident logs)

03 SKILL.md 正文

📊 Skill: samm-maturity

Context

OWASP SAMM (Software Assurance Maturity Model) provides a structured framework for assessing software security practices across five business functions, each with three security practices, totaling 15 streams. Maturity is scored 0–3 per stream (Level 0 = ad-hoc/absent, Level 1 = performed, Level 2 = quantified/managed, Level 3 = optimized).

samm-maturity standardizes engagement-based security posture assessment by scoring all 15 streams against evidence, computing per-function and overall maturity, and generating a prioritized improvement roadmap.

When to Use

  • Performing baseline security posture assessment during {{PROJECT_PHASE_GATES}}.
  • Evaluating security program maturity before recommending tooling or process investments.
  • Tracking security improvement progress across multiple engagement cycles.
  • Generating stakeholder-ready maturity reports for governance, risk, and compliance (GRC) reviews.

Execution Steps

  1. Scope the Engagement & Gather Evidence

    • Define assessment scope: organization, business unit, or product line.
    • Collect evidence per stream: policy documents, architecture diagrams, CI/CD configs, scan reports, incident logs, training records.
    • Interview stakeholders: security leadership, engineering managers, DevOps owners, incident responders.
  2. Score Each of the 15 Streams (0–3) For each stream, assign a maturity level with one-line evidence citation:

    Governance

    • Strategy & Metrics (G1): Security strategy, objectives, metrics program
    • Policy & Compliance (G2): Security policies, regulatory compliance tracking
    • Education & Guidance (G3): Security training, guidance docs, secure development resources

    Design

    • Threat Assessment (D1): Threat modeling, attack surface analysis
    • Secure Architecture (D2): Secure design patterns, reference architectures
    • Security Requirements (D3): Security requirements definition, acceptance criteria

    Implementation

    • Secure Build (I1): Secure build practices, dependency management, SAST/DAST
    • Secure Deployment (I2): Secure deployment automation, environment hardening
    • Defect Management (I3): Vulnerability tracking, remediation SLAs

    Verification

    • Implementation Verification (V1): Code review, manual security testing
    • Security Testing (V2): Automated security testing in CI/CD
    • Requirements-driven Testing (V3): Test coverage of security requirements

    Operations

    • Incident Detection (O1): Security monitoring, alerting, log analysis
    • Incident Response (O2): Incident response playbook, forensics, communication
    • Operational Management (O3): Configuration management, patch management, access control

    Maturity Level Rubric:

    • Level 0: Ad-hoc or absent (no documented process)
    • Level 1: Performed (process exists but not consistent across org)
    • Level 2: Quantified/managed (measured, tracked, some automation)
    • Level 3: Optimized (fully automated, quantitative metrics, continuous improvement)
  3. Compute Maturity Scores

    • Calculate per-function average: (sum of 3 stream scores) / 3
    • Calculate overall maturity score: (sum of all 15 stream scores) / 15
    • Round scores to one decimal place (e.g., 1.7, 2.3)
  4. Build Target-State Roadmap For each stream below target maturity (typically Level 2 or 3):

    • Identify current score and target score
    • Define one concrete improvement action (S/M/L size)
    • Tie action to the next maturity level step
    • Prioritize by risk impact and implementation effort
  5. Generate Output Artifact

    • Compile markdown assessment report with summary table and roadmap
    • Save to docs/security/samm-assessment-<YYYY-MM-DD>.md

Output Format

# OWASP SAMM Maturity Assessment

**Organization**: [Organization Name]
**Assessment Date**: 2026-08-24
**Evaluator**: [Evaluator Name]

## Executive Summary

**Overall Maturity**: 1.7 / 3.0

| Business Function | Stream Scores | Average |
|---|---|---|
| Governance | G1: 2, G2: 1, G3: 2 | 1.7 |
| Design | D1: 1, D2: 1, D3: 1 | 1.0 |
| Implementation | I1: 2, I2: 2, I3: 1 | 1.7 |
| Verification | V1: 1, V2: 1, V3: 0 | 0.7 |
| Operations | O1: 2, O2: 2, O3: 2 | 2.0 |

## Detailed Stream Scores

### Governance (1.7 / 3.0)
- **G1 Strategy & Metrics**: Level 2 — Metrics program established with quarterly KPI reviews (evidence: `docs/security/metrics-2026-Q2.md`)
- **G2 Policy & Compliance**: Level 1 — Policies exist but compliance tracking is manual (evidence: `docs/policies/`)
- **G3 Education & Guidance**: Level 2 — Annual security training plus secure coding guidelines (evidence: `docs/training/`)

### Design (1.0 / 3.0)
- **D1 Threat Assessment**: Level 1 — Ad-hoc threat modeling, no standardized process
- **D2 Secure Architecture**: Level 1 — Secure patterns documented but not enforced
- **D3 Security Requirements**: Level 1 — Requirements defined per project, no standard catalog

### Implementation (1.7 / 3.0)
- **I1 Secure Build**: Level 2 — SAST in CI/CD, dependency scanning enabled (evidence: `.github/workflows/`)
- **I2 Secure Deployment**: Level 2 — Infrastructure as code with security checks (evidence: `terraform/`)
- **I3 Defect Management**: Level 1 — Vulnerability tracking in Jira, inconsistent SLAs

### Verification (0.7 / 3.0)
- **V1 Implementation Verification**: Level 1 — Manual code reviews, no standardized checklist
- **V2 Security Testing**: Level 1 — DAST on production only, no pre-production testing
- **V3 Requirements-driven Testing**: Level 0 — No security test coverage tracking

### Operations (2.0 / 3.0)
- **O1 Incident Detection**: Level 2 — SIEM deployed with alert rules (evidence: `docs/siem-rules.md`)
- **O2 Incident Response**: Level 2 — Documented playbook, quarterly drills (evidence: `docs/ir-playbook.md`)
- **O3 Operational Management**: Level 2 — Automated patch management, quarterly access reviews

## Prioritized Improvement Roadmap

| Stream | Current | Target | Action | Size | Priority |
|---|---|---|---|---|---|
| V3 Requirements-driven Testing | 0 | 2 | Implement security test coverage tracking in CI/CD | M | High |
| D1 Threat Assessment | 1 | 2 | Standardize threat modeling process with template | S | High |
| I3 Defect Management | 1 | 2 | Define vulnerability SLAs and automate tracking | M | High |
| D2 Secure Architecture | 1 | 2 | Enforce secure patterns via architecture review gate | M | Medium |
| V2 Security Testing | 1 | 2 | Integrate DAST into pre-production pipeline | M | Medium |

**Size Legend**: S (1-2 weeks, <5 person-days), M (3-8 weeks, 5-20 person-days), L (9+ weeks, >20 person-days)

Co-develop Integration Note

This file is the hub-doctrine canonical for the co-security assurance suite (spec 2026-11-01-hub-consolidation-design, Phase 1 batch C). Co-security engagements run it as written. A co-develop pipeline instance derives from this doctrine by substituting the bindings below — the derivation never edits frontmatter: owner: pm stays single-valued here, and the co-develop ownership binding lives in this note only.

Parameterization — unfilled placeholders resolve to the co-security doctrine defaults; a co-develop derivation substitutes the bindings:

Placeholder Instance Unfilled default (co-security) Co-develop binding
{{PROJECT_PHASE_GATES}} When to Use, first bullet Phase 1 security engagements Phase 0 security baseline scans

Co-develop bindings beyond placeholders:

  • Ownership: co-develop renders this skill as co-owned pm, security-monitor — the Phase 0 security baseline scans it feeds are security-monitor scans. Frontmatter here stays owner: pm.
  • Delivery: the co-develop fork carries scope: common; this canonical keeps scope: co-security (the engagement label stays co-security for hub readers).

Related Skills

  • stride-threat-matrix — feeds Design/Verification evidence (D1, D2, V1, V2) by identifying structural threats and risk ratings.
  • sarif-exporter — feeds Implementation/Verification evidence (I1, I3, V2) by providing vulnerability findings and scan results in standardized format.
  • security-scan — produces SAST/DAST/secret detection results ingested into I1, I3, V2 scoring.

04 同仓库更多 Skills

5throck
5throck/ai-workspace-standards

i18n-layout

Text layout and encoding guidance: character encoding (UTF-8, legacy Korean code pages, BOM hazards), line endings, RTL/bidi handling, and script-specific font selection (Hangul focus). Use when: encoding is corrupted or mojibake appears, an RTL locale must render correctly, fonts must be chosen for a script, or CRLF/BOM issues surface.

★ 25
5throck
5throck/ai-workspace-standards

i18n-layout

Text layout and encoding guidance: character encoding (UTF-8, legacy Korean code pages, BOM hazards), line endings, RTL/bidi handling, and script-specific font selection (Hangul focus). Use when: encoding is corrupted or mojibake appears, an RTL locale must render correctly, fonts must be chosen for a script, or CRLF/BOM issues surface.

★ 25
5throck
5throck/ai-workspace-standards

i18n-layout

Text layout and encoding guidance: character encoding (UTF-8, legacy Korean code pages, BOM hazards), line endings, RTL/bidi handling, and script-specific font selection (Hangul focus). Use when: encoding is corrupted or mojibake appears, an RTL locale must render correctly, fonts must be chosen for a script, or CRLF/BOM issues surface.

★ 25
5throck
5throck/ai-workspace-standards

i18n-layout

Text layout and encoding guidance: character encoding (UTF-8, legacy Korean code pages, BOM hazards), line endings, RTL/bidi handling, and script-specific font selection (Hangul focus). Use when: encoding is corrupted or mojibake appears, an RTL locale must render correctly, fonts must be chosen for a script, or CRLF/BOM issues surface.

★ 25
5throck
5throck/ai-workspace-standards

i18n-layout

Text layout and encoding guidance: character encoding (UTF-8, legacy Korean code pages, BOM hazards), line endings, RTL/bidi handling, and script-specific font selection (Hangul focus). Use when: encoding is corrupted or mojibake appears, an RTL locale must render correctly, fonts must be chosen for a script, or CRLF/BOM issues surface.

★ 25
5throck
5throck/ai-workspace-standards

i18n-layout

Text layout and encoding guidance: character encoding (UTF-8, legacy Korean code pages, BOM hazards), line endings, RTL/bidi handling, and script-specific font selection (Hangul focus). Use when: encoding is corrupted or mojibake appears, an RTL locale must render correctly, fonts must be chosen for a script, or CRLF/BOM issues surface.

★ 25