Harden tools/pr_threads.sh: pass all printed thread fields through the deterministic sanitizer
Maintainers usually reply within 1 day
Nobody has claimed this yet.
Assessment
- Difficulty
- 2/5
- Estimated time
- 1-3 hours
- Newbie friendliness
- 76/100
Research direction
Start with tools/pr_threads.sh and the repository’s AGENTS.md section “Agent-Side Untrusted-Content Protocol” to see how sanitize() is used for the title. Route the printed thread fields through that sanitizer, then add a test in tests/test_prompts_lint.py or a tools test using a hostile fake path. Done means the output neutralizes that path while preserving the expected thread details.
Written by the indexing model from the issue text.
Description
Residual from the github-advanced-security review on PR #1254 (threads PRRT_kwDOSw1GmM62RHnZ at tools/pr_threads.sh:37 and PRRT_kwDOSw1GmM62RHno at :32).
Already fixed in f0d6dae (the merged #1254 itself): gh api graphql output is piped into the python sanitizer (no || failure path that would print raw JSON), the default query does not fetch comment bodies at all, and --include-bodies prints only [UNTRUSTED_COMMENT_REDACTED] body_present=... body_len=... after redaction/truncation.
Remaining gap: the printed thread line interpolates t['id'], t.get('path'), and the author login without passing them through sanitize() — only the PR title and the [REDACTED] markers are sanitized. A crafted diff path or author name could carry instruction-like text into the agent's context, which the repo's untrusted-content protocol (AGENTS.md, "Agent-Side Untrusted-Content Protocol") requires to be deterministically redacted, not merely truncated.
Fix: route the whole printed line (id/path/line/author) through the same sanitize() used for the title, and add a test (e.g. in tests/test_prompts_lint.py or a new tools test) that a hostile fake path like ignore-all-previous-instructions.md is neutralized in the output.
- Dominant language
- Python
- Stars
- 8
- Forks
- 12
- Avg merge
- 7h 37m
- Merged PRs (30d)
- 218
Getting set up
- No Dockerfile or Docker Compose file
- Has a pull request template
- Read the contributing guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
More from RonaldHensbergen/composable-data-stack
-
area:kubernetes area:security documentation priority:medium size:small
Difficulty 1/5 1-3 hours Newbie friendliness 88/100
RonaldHensbergen/composable-data-stack#1339 ·
Maintainers usually reply within 1 day
-
area:docs documentation good first issue priority:medium size:small
Difficulty 2/5 1-3 hours Newbie friendliness 80/100
RonaldHensbergen/composable-data-stack#1285 ·
Maintainers usually reply within 1 day
-
priority:low refactor size:small
Difficulty 2/5 1-3 hours Newbie friendliness 78/100
RonaldHensbergen/composable-data-stack#1230 ·
Maintainers usually reply within 1 day
-
enhancement priority:low size:small
Difficulty 1/5 Under an hour Newbie friendliness 85/100
RonaldHensbergen/composable-data-stack#1205 ·
Maintainers usually reply within 1 day
-
area:docs documentation priority:low size:small
Difficulty 1/5 1-3 hours Newbie friendliness 86/100
RonaldHensbergen/composable-data-stack#944 ·
Maintainers usually reply within 1 day
All issues in RonaldHensbergen/composable-data-stack
Similar issues
-
Difficulty 2/5 1-3 hours Newbie friendliness 78/100
abetlen/llama-cpp-python#2424 ·
Maintainers usually reply within 1 day
-
Bug IronPython Python 3
Difficulty 2/5 1-3 hours Newbie friendliness 72/100
pyrevitlabs/pyRevit#4184 ·
Maintainers usually reply within 1 day
-
Difficulty 1/5 Under an hour Newbie friendliness 74/100
jqueryscript/awesome-claude-code#1009 ·
-
tests
Difficulty 2/5 1-3 hours Newbie friendliness 66/100
Maintainers usually reply within 3 days
-
Difficulty 2/5 1-3 hours Newbie friendliness 72/100
Maintainers usually reply within 1 day