You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
refactor: move templates into src/data/ directory (#235)
Move pipeline templates from top-level templates/ into src/data/
alongside ecosystem_domains.json, keeping injected content separate
from source code. Update all include paths, documentation, workflow
path filters, and test references.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Copy file name to clipboardExpand all lines: .github/workflows/red-team-security.md
+10-8Lines changed: 10 additions & 8 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -48,7 +48,7 @@ Use a **round-robin approach** across these categories. Pick up where the last r
48
48
49
49
### Category A: Input Sanitization & Injection
50
50
51
-
Audit `src/sanitize.rs`, `src/compile/types.rs`, `src/compile/common.rs`, and `src/compile/standalone.rs` for:
51
+
Audit `src/sanitize.rs`, `src/compile/types.rs`, `src/compile/common.rs`, `src/compile/standalone.rs`, and `src/compile/onees.rs` for:
52
52
53
53
-**Template injection**: Can a malicious `name`, `description`, or other front matter field inject ADO template expressions (dollar-double-brace syntax) into generated YAML that Azure DevOps evaluates?
54
54
-**YAML deserialization**: Can crafted front matter trigger unexpected serde_yaml behavior (anchors, aliases, merge keys, billion-laughs)?
@@ -61,8 +61,9 @@ Focus files:
61
61
cat src/sanitize.rs
62
62
cat src/compile/common.rs
63
63
cat src/compile/standalone.rs
64
-
grep -n 'format!' src/compile/standalone.rs | head -40
Audit `src/compile/common.rs`, `Cargo.toml`, and `templates/base.yml` for:
140
+
Audit `src/compile/common.rs`, `Cargo.toml`, `src/data/base.yml`, and `src/data/1es-base.yml` for:
139
141
140
142
-**Binary integrity**: Are the `ado-aw`, AWF, and MCPG binaries downloaded with proper checksum verification? Can the checksums file itself be tampered with?
141
143
-**Docker image pinning**: Is the MCPG Docker image pinned by digest, or only by tag? Tag-only pinning allows image replacement attacks.
@@ -540,8 +541,8 @@ When using `target: 1es`, the pipeline will extend `1es/1ES.Unofficial.PipelineT
540
541
541
542
The compiler transforms the input into valid Azure DevOps pipeline YAML based on the target platform:
542
543
543
-
- **Standalone**: Uses `templates/base.yml`
544
-
- **1ES**: Uses `templates/1es-base.yml`
544
+
- **Standalone**: Uses `src/data/base.yml`
545
+
- **1ES**: Uses `src/data/1es-base.yml`
545
546
546
547
Explicit markings are embedded in these templates that the compiler is allowed to replace e.g. `{{ copilot_params }}` denotes parameters which are passed to the copilot command line tool. The compiler should not replace sections denoted by `${{ some content }}`. What follows is a mapping of markings to responsibilities (primarily for the standalone template).
547
548
@@ -831,7 +832,7 @@ Example output:
831
832
832
833
## {{ threat_analysis_prompt }}
833
834
834
-
Should be replaced with the embedded threat detection analysis prompt from `templates/threat-analysis.md`. This prompt template includes markers for `{{ source_path }}`, `{{ agent_name }}`, `{{ agent_description }}`, and `{{ working_directory }}` which are replaced during compilation.
835
+
Should be replaced with the embedded threat detection analysis prompt from `src/data/threat-analysis.md`. This prompt template includes markers for `{{ source_path }}`, `{{ agent_name }}`, `{{ agent_description }}`, and `{{ working_directory }}` which are replaced during compilation.
835
836
836
837
The threat analysis prompt instructs the security analysis agent to check for:
0 commit comments