python-packaging-git-audit

Name: python-packaging-git-audit
Author: opendatahub-io

36stars74forks

How to install python-packaging-git-audit?

opendatahub-io/ai-helpers/python-packaging-git-audit

$npx -y skills add opendatahub-io/ai-helpers --skill python-packaging-git-audit

Installs into the current project.

Summary

Inspect recent git history of a Python package repository for suspicious commits touching supply-chain-sensitive files, then triage findings with AI reasoning to produce a structured risk report section.

Files · 1

View on GitHub

SKILL.md

1	# Python Packaging Git Audit
2
3	Inspects the recent git history of a Python package repository for commits
4	that modify supply-chain-sensitive files (setup.py, CI configs, .pth files,
5	etc.) and contain suspicious patterns. Produces a self-contained "Git History
6	Analysis" report section with triaged findings and a risk assessment.
7
8	## Inputs
9
10	- repo_path (required): Local filesystem path to a git repository
11	- output_file (optional): Write the report section to this file path instead of
12	returning it inline. The first line of the file must be `RISK_RATING:<value>` so
13	the orchestrator can parse it without reading the full report.
14
15	## Step 1: Enumerate Commits
16
17	Determine the number of commits available and scan up to the last 50:
18
19	```bash
20	git -C <repo-path> log -50 --format='%H'
21	```
22
23	## Step 2: Filter Sensitive Files
24
25	From those commits, identify ones that modify supply-chain-sensitive files:
26
27	- `setup.py`, `setup.cfg`, `pyproject.toml`, `MANIFEST.in`
28	- `.github/workflows/*.yml`, `.gitlab-ci.yml`
29	- Any `.pth` files
30	- `__init__.py` files at package root
31
32	```bash
33	git -C <repo-path> log -50 --diff-filter=ACMR --name-only --format='COMMIT:%H\|%aI\|%ae' -- \
34	setup.py setup.cfg pyproject.toml MANIFEST.in \
35	'.github/workflows/*.yml' .gitlab-ci.yml \
36	'.pth' '/__init__.py'
37	```
38
39	## Step 3: Search for Suspicious Patterns
40
41	For each flagged commit, extract the diff and search for suspicious patterns:
42
43	- Code execution: `eval(`, `exec(`, `compile(`
44	- Process spawning: `subprocess`, `os.system`, `os.popen`
45	- Encoding/serialization: `base64`, `marshal`, `pickle`
46	- Network access: `socket`, `urllib`, `requests.get`, `httpx`
47	- Native code: `ctypes`, `cffi`
48	- Embedded URLs: `http://`, `https://`
49
50	```bash
51	git -C "<repo-path>" show --format= -m --first-parent "<commit>" -- "<file>" \| \
52	grep -nE 'eval\(\|exec\(\|compile\(\|subprocess\|os\.system\|os\.popen\|base64\|marshal\|pickle\|socket\|urllib\|requests\.get\|httpx\|ctypes\|cffi\|https?://'
53	```
54
55	Collect per finding: commit hash, author, date, file modified, and matching patterns.
56
57	## Step 4: Triage
58
59	Review git history findings with AI reasoning. There are no deterministic rules
60	for this phase — each finding requires contextual judgment.
61
62	For each flagged commit, consider:
63
64	- Is this a normal dependency version bump or configuration change?
65	- Does the pattern match the package's stated purpose?
66	- Is the author a known maintainer with a history of contributions?
67	- Does the change introduce new capabilities that seem out of scope?
68
69	Flag commits that introduced suspicious patterns into sensitive files. Assign
70	a verdict to each:
71
72	- Likely legitimate — change is a normal maintenance or dependency update
73	- Suspicious — change introduces unexpected capabilities or patterns
74	- Critical — change strongly indicates malicious intent (e.g., encoded payload in install hook)
75
76	## Output Format
77
78	Produce the following markdown section:
79
80	```markdown
81	## Git History Analysis (Last 50 Commits)
82
83	Commits touching sensitive files: {N}
84
85	\| Commit \| Date \| Author \| File \| Patterns Found \|
86	\|--------\|------\|--------\|------\|----------------\|
87	\| abc1234 \| 2026-03-15 \| user@example.com \| setup.py \| subprocess, os.system \|
88
89	AI Assessment: {Brief narrative on whether the git history changes look normal or concerning, with reasoning}
90	```
91
92	The risk_rating for this phase is one of:
93
94	- no_issues — No sensitive files modified in recent history
95	- low_risk — All findings classified as "likely legitimate"
96	- needs_review — One or more findings classified as "suspicious"
97	- critical — One or more findings classified as "critical"
98
99	If `output_file` is provided, write the file with the first line as
100	`RISK_RATING:<value>` followed by a blank line and then the markdown section
101	above. If `output_file` is not provided, return the report section inline.
102
103	## Error Handling
104
105	\| Scenario \| Behavior \|
106	\|----------\|----------\|
107	\| Path is not a git repository \| Report "not a git repository", risk_rating = needs_review \|
108	\| Fewer than 50 commits available \| Scan all available commits, note count in report \|
109	\| No sensitive files modified in history \| Report "no sensitive files found", risk_rating = no_issues \|

People who liked this also viewed

6 picks

Type

python-packaging-git-audit

36stars74forks

View on GitHub

How to install python-packaging-git-audit?

opendatahub-io/ai-helpers/python-packaging-git-audit

$npx -y skills add opendatahub-io/ai-helpers --skill python-packaging-git-audit

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Python Packaging Git Audit
2
3	Inspects the recent git history of a Python package repository for commits
4	that modify supply-chain-sensitive files (setup.py, CI configs, .pth files,
5	etc.) and contain suspicious patterns. Produces a self-contained "Git History
6	Analysis" report section with triaged findings and a risk assessment.
7
8	## Inputs
9
10	- repo_path (required): Local filesystem path to a git repository
11	- output_file (optional): Write the report section to this file path instead of
12	returning it inline. The first line of the file must be `RISK_RATING:<value>` so
13	the orchestrator can parse it without reading the full report.
14
15	## Step 1: Enumerate Commits
16
17	Determine the number of commits available and scan up to the last 50:
18
19	```bash
20	git -C <repo-path> log -50 --format='%H'
21	```
22
23	## Step 2: Filter Sensitive Files
24
25	From those commits, identify ones that modify supply-chain-sensitive files:
26
27	- `setup.py`, `setup.cfg`, `pyproject.toml`, `MANIFEST.in`
28	- `.github/workflows/*.yml`, `.gitlab-ci.yml`
29	- Any `.pth` files
30	- `__init__.py` files at package root
31
32	```bash
33	git -C <repo-path> log -50 --diff-filter=ACMR --name-only --format='COMMIT:%H\|%aI\|%ae' -- \
34	setup.py setup.cfg pyproject.toml MANIFEST.in \
35	'.github/workflows/*.yml' .gitlab-ci.yml \
36	'.pth' '/__init__.py'
37	```
38
39	## Step 3: Search for Suspicious Patterns
40
41	For each flagged commit, extract the diff and search for suspicious patterns:
42
43	- Code execution: `eval(`, `exec(`, `compile(`
44	- Process spawning: `subprocess`, `os.system`, `os.popen`
45	- Encoding/serialization: `base64`, `marshal`, `pickle`
46	- Network access: `socket`, `urllib`, `requests.get`, `httpx`
47	- Native code: `ctypes`, `cffi`
48	- Embedded URLs: `http://`, `https://`
49
50	```bash
51	git -C "<repo-path>" show --format= -m --first-parent "<commit>" -- "<file>" \| \
52	grep -nE 'eval\(\|exec\(\|compile\(\|subprocess\|os\.system\|os\.popen\|base64\|marshal\|pickle\|socket\|urllib\|requests\.get\|httpx\|ctypes\|cffi\|https?://'
53	```
54
55	Collect per finding: commit hash, author, date, file modified, and matching patterns.
56
57	## Step 4: Triage
58
59	Review git history findings with AI reasoning. There are no deterministic rules
60	for this phase — each finding requires contextual judgment.
61
62	For each flagged commit, consider:
63
64	- Is this a normal dependency version bump or configuration change?
65	- Does the pattern match the package's stated purpose?
66	- Is the author a known maintainer with a history of contributions?
67	- Does the change introduce new capabilities that seem out of scope?
68
69	Flag commits that introduced suspicious patterns into sensitive files. Assign
70	a verdict to each:
71
72	- Likely legitimate — change is a normal maintenance or dependency update
73	- Suspicious — change introduces unexpected capabilities or patterns
74	- Critical — change strongly indicates malicious intent (e.g., encoded payload in install hook)
75
76	## Output Format
77
78	Produce the following markdown section:
79
80	```markdown
81	## Git History Analysis (Last 50 Commits)
82
83	Commits touching sensitive files: {N}
84
85	\| Commit \| Date \| Author \| File \| Patterns Found \|
86	\|--------\|------\|--------\|------\|----------------\|
87	\| abc1234 \| 2026-03-15 \| user@example.com \| setup.py \| subprocess, os.system \|
88
89	AI Assessment: {Brief narrative on whether the git history changes look normal or concerning, with reasoning}
90	```
91
92	The risk_rating for this phase is one of:
93
94	- no_issues — No sensitive files modified in recent history
95	- low_risk — All findings classified as "likely legitimate"
96	- needs_review — One or more findings classified as "suspicious"
97	- critical — One or more findings classified as "critical"
98
99	If `output_file` is provided, write the file with the first line as
100	`RISK_RATING:<value>` followed by a blank line and then the markdown section
101	above. If `output_file` is not provided, return the report section inline.
102
103	## Error Handling
104
105	\| Scenario \| Behavior \|
106	\|----------\|----------\|
107	\| Path is not a git repository \| Report "not a git repository", risk_rating = needs_review \|
108	\| Fewer than 50 commits available \| Scan all available commits, note count in report \|
109	\| No sensitive files modified in history \| Report "no sensitive files found", risk_rating = no_issues \|

People who liked this also viewed

6 picks

Type