Skip to content

Commit 610a0be

Browse files
dgerogclaude
andcommitted
Update docs: entity-level report commands, compliance guidance
- Rewrite reports.md with all 4 report levels (project, org, assessment, experiment) - Document report content, options, and compliance use cases - Update commands.md reference: remove deprecated hb report, add entity-level commands Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
1 parent 7b3684c commit 610a0be

2 files changed

Lines changed: 72 additions & 15 deletions

File tree

docs/docs/reference/commands.md

Lines changed: 4 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -58,9 +58,12 @@ Complete reference of all available commands, organized by category.
5858
| `hb findings` | List persistent vulnerability findings |
5959
| `hb findings update <id>` | Update finding status or severity |
6060
| `hb findings assign <id>` | Assign finding to a team member |
61-
| `hb report` | Generate HTML security report (project, org, or assessment) |
6261
| `hb assessments` | List past security assessments |
6362
| `hb assessments show <id>` | View assessment detail (posture before/after, drift, test count) |
63+
| `hb assessments report <id>` | Generate assessment HTML report with full test logs |
64+
| `hb projects report` | Generate project HTML security report |
65+
| `hb orgs report` | Generate organisation-wide HTML report |
66+
| `hb experiments report <id>` | Generate experiment HTML report with methodology context |
6467

6568
## Security
6669

docs/docs/testing/reports.md

Lines changed: 68 additions & 14 deletions
Original file line numberDiff line numberDiff line change
@@ -1,33 +1,87 @@
11
# Reports
22

3-
Generate HTML security assessment reports for projects, assessments, or the entire organisation.
3+
Generate branded HTML security reports at four levels. Reports are generated by the backend and downloaded locally — they open automatically in your browser.
44

5-
## Project Report
5+
All reports include:
6+
7+
- **Methodology section** — testing approach, posture scoring, continuous monitoring
8+
- **Technology disclaimer** — LLM stochastic nature, limitations
9+
- **Legal notice** — copyright, confidentiality, no tampering
10+
- **Print-ready CSS** — use browser "Print to PDF" for compliance submissions
11+
12+
## Report Levels
13+
14+
### Project Report
15+
16+
The standing security posture of an agent — findings, threat landscape, monitoring status, and assessment history.
617

718
```bash
8-
# Generate report for current project
9-
hb report
19+
# Current project
20+
hb projects report
21+
22+
# Save without opening browser
23+
hb projects report --no-open
1024

11-
# Save to file
12-
hb report --output report.html
25+
# Custom output path
26+
hb projects report -o ./reports/q1-security.html
1327
```
1428

15-
## Organisation Report
29+
**Includes:** Agent scope (permitted/restricted operations), posture donuts (overall + security + quality), findings with severity and threat class, threat landscape, assessment history (last 90 days), human feedback audit summary.
30+
31+
### Organisation Report
32+
33+
Executive overview across all projects in the organisation.
1634

1735
```bash
18-
# Org-level report (all projects + inventory)
19-
hb report --org
36+
hb orgs report
37+
hb orgs report -o org-report.html
2038
```
2139

22-
## Assessment Report
40+
**Includes:** Organisation posture donuts, findings summary with severity bar, all projects with grade, score, last assessed date, and monitoring status.
41+
42+
### Assessment Report
43+
44+
What happened in a specific test run — results, findings, and full conversation evidence.
2345

2446
```bash
25-
# Report for a specific assessment
26-
hb report --assessment <assessment-id>
47+
# Get assessment ID from list
48+
hb assessments
49+
50+
# Generate report
51+
hb assessments report <assessment-id>
52+
hb assessments report abc123 -o assessment.html
2753
```
2854

29-
## JSON Output
55+
**Includes:** Overview (tests, pass rate, status), test suite (engines, level, language), posture before/after donuts, findings, and a full appendix of every test conversation with verdict, severity, explanation, and multi-turn dialogue.
56+
57+
### Experiment Report
58+
59+
Deep dive into a single test engine's run, with orchestrator-specific methodology context.
3060

3161
```bash
32-
hb report --json
62+
# Get experiment ID from list
63+
hb experiments list
64+
65+
# Generate report
66+
hb experiments report <experiment-id>
67+
hb experiments report abc123 -o experiment.html
3368
```
69+
70+
**Includes:** Orchestrator-specific context (OWASP methodology for adversarial, QA evaluation dimensions for behavioral), metrics (TPI, reliability, pass rate), vulnerabilities identified, and full conversation appendix with feedback badges.
71+
72+
## Options
73+
74+
| Option | Description |
75+
|--------|-------------|
76+
| `-o`, `--output PATH` | Custom output file path |
77+
| `--no-open` | Save file without opening in browser |
78+
79+
## For Compliance
80+
81+
Reports are designed for submission to auditors and compliance frameworks including DORA, PCI-DSS, ISO/IEC 42001, NIS2, and the EU AI Act.
82+
83+
- **Project reports** prove ongoing monitoring and scope definition
84+
- **Assessment reports** provide test evidence with full conversation logs
85+
- **Experiment reports** document specific testing methodology
86+
87+
Use browser "Print to PDF" to generate PDF versions suitable for formal submissions.

0 commit comments

Comments
 (0)