Add docs cherry pick command#48
Open
kquinn1204 wants to merge 9 commits intoredhat-documentation:mainfrom
Open
Add docs cherry pick command#48kquinn1204 wants to merge 9 commits intoredhat-documentation:mainfrom
kquinn1204 wants to merge 9 commits intoredhat-documentation:mainfrom
Conversation
|
Note Reviews pausedIt looks like this branch is under active development. To avoid overwhelming you with review comments due to an influx of new commits, CodeRabbit has automatically paused this review. You can configure this behavior by changing the Use the following commands to manage reviews:
Use the checkboxes below for quick actions:
WalkthroughAdds docs and scripts implementing a documentation cherry-pick backport workflow: a new command doc, a branch-audit skill (with deep-audit), two executable Bash scripts for per-branch existence and deep applicability checks, a README Vale reminder for AsciiDocDITA, and a plugin manifest version bump. Changes
Sequence Diagram(s)sequenceDiagram
participant User
participant CLI as docs-cherry-pick
participant PRReader as git_pr_reader/gh
participant BranchAudit as docs-branch-audit (branch_audit.sh)
participant DeepAudit as deep_audit.sh
participant Git as Git (local/remote)
participant Agent as ConflictResolver
User->>CLI: invoke with source + --target(s) + options
CLI->>PRReader: fetch PR/commit metadata and file list
PRReader-->>CLI: return changed .adoc paths and diff
CLI->>BranchAudit: audit file presence across target branches
BranchAudit-->>CLI: include/exclude lists per branch
alt deep mode
CLI->>DeepAudit: run per-branch deep comparisons
DeepAudit-->>CLI: per-file confidence + issues
end
CLI->>User: present included/excluded files and confirmation
User-->>CLI: approve
loop per target branch
CLI->>Git: create backport branch
CLI->>Git: cherry-pick commit(s) (no-commit)
alt conflicts detected
CLI->>Agent: attempt automated per-file resolution
Agent-->>CLI: resolution or mark for review
CLI->>User: request manual resolution for REVIEW items
end
CLI->>Git: commit staged changes with structured message
opt push
CLI->>Git: push branch and prepare PR descriptor
end
end
Estimated code review effort🎯 4 (Complex) | ⏱️ ~45 minutes 🚥 Pre-merge checks | ✅ 2 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (2 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 2
🧹 Nitpick comments (9)
plugins/docs-tools/skills/docs-branch-audit/SKILL.md (1)
61-75: Add language specifier to fenced code blocks.The linter flags these output examples as missing a language specifier. Use
textorconsoleto satisfy MD040:-``` +=== Branch Audit: enterprise-4.17 === +```text +=== Branch Audit: enterprise-4.17 ===Also applies to: 79-99
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@plugins/docs-tools/skills/docs-branch-audit/SKILL.md` around lines 61 - 75, The fenced code blocks showing the branch audit output (e.g., the block starting with "=== Branch Audit: enterprise-4.17 ===" and other similar audit/example blocks) are missing a language specifier; update each triple-backtick fence to include a language such as text or console (for example replace ``` with ```text) so the linter MD040 is satisfied for the blocks referenced around the "Branch Audit" examples.plugins/docs-tools/skills/docs-branch-audit/scripts/deep_audit.sh (1)
141-157:git apply --checkmay not work as expected without proper worktree context.The patch is extracted from a unified diff and
git apply --check --3wayis run, but this operates on the current worktree state, not the target branch content. For the check to be meaningful, the worktree would need to reflect the target branch version of the file.Consider either:
- Documenting this limitation (the check is best-effort)
- Using
git apply --check --3waywith an explicit index built from the target refGiven this is a heuristic audit tool, this may be acceptable as-is with a comment noting the limitation.
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@plugins/docs-tools/skills/docs-branch-audit/scripts/deep_audit.sh` around lines 141 - 157, The current git apply --check --3way invocation (when checking "$FILE_PATCH") operates against the current worktree, not the target branch snapshot; update the deep_audit.sh logic around the FILE_PATCH check to either (A) add a clear comment documenting this limitation as a best-effort heuristic, or (B) implement a safer check by building an explicit temporary index/tree from the target ref and running git apply --check --3way against that index (use the same FILE_PATCH and preserve setting of CONFIDENCE and ISSUES), and ensure the CONFIDENCE and ISSUES handling (patch-conflict / patch-mismatch / patch: applies cleanly) remains unchanged.plugins/docs-tools/commands/docs-cherry-pick.md (4)
614-617: Use fully qualified skill names in references.Per coding guidelines, skill references should use the
plugin:skillformat:## Related -- **docs-branch-audit skill**: The underlying file existence and content comparison check used by this command -- **git-pr-reader skill**: PR/MR file listing and diff extraction used for source analysis +- **docs-tools:docs-branch-audit**: The underlying file existence and content comparison check used by this command +- **docs-tools:git-pr-reader**: PR/MR file listing and diff extraction used for source analysis🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@plugins/docs-tools/commands/docs-cherry-pick.md` around lines 614 - 617, Update the two Related entries to use fully qualified skill names in the plugin:skill format; replace "docs-branch-audit skill" with "plugin:docs-branch-audit" and "git-pr-reader skill" with "plugin:git-pr-reader" (preserve the descriptive text/links and formatting). Ensure the markdown list under "## Related" uses the exact plugin:skill identifiers so they follow the project's naming guideline.
1-6: Missing recommended documentation sections.The pipeline lint indicates this command documentation is missing recommended sections:
## Name,## Synopsis,## Description, and## Implementation. Consider adding these sections for consistency with other command docs.📝 Suggested structure
--- description: Intelligently cherry-pick documentation changes to enterprise branches... argument-hint: <pr-url|--commit sha> --target <branch> [--dry-run] allowed-tools: Read, Write, Glob, Grep, Edit, Bash, Skill, AskUserQuestion, Agent --- ## Name docs-cherry-pick — Backport documentation changes to enterprise branches ## Synopsis/docs-tools:docs-cherry-pick --target [--dry-run] [--deep] [--no-push]
/docs-tools:docs-cherry-pick --commit --target [--ticket ]## Description Intelligently backport documentation changes from a PR or commit... ## Implementation <!-- existing workflow content -->🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@plugins/docs-tools/commands/docs-cherry-pick.md` around lines 1 - 6, Add the recommended documentation sections to the command doc by inserting top-level headings "## Name", "## Synopsis", "## Description", and "## Implementation" after the existing YAML frontmatter in plugins/docs-tools/commands/docs-cherry-pick.md; populate "## Name" with a short one-line command description (e.g., docs-cherry-pick — Backport documentation changes to enterprise branches), "## Synopsis" with example CLI usages (showing both PR URL and --commit patterns and common flags like --target, --dry-run, --deep, --no-push, --ticket), "## Description" with a concise paragraph describing the command behavior (intelligently cherry-picks/backports docs changes and excludes files not present on target releases), and "## Implementation" with a note or the existing workflow details that explain how the command performs the backporting so the file matches the suggested structure.
105-105: Samegrep -oPportability issue.This line uses
grep -oPwhich is not available on macOS's default grep. Use the same sed-based alternative suggested forbranch_audit.sh.🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@plugins/docs-tools/commands/docs-cherry-pick.md` at line 105, The PR_NUMBER extraction uses GNU grep's -oP which is not portable; update the assignment that sets PR_NUMBER from SOURCE (the line referencing PR_NUMBER and SOURCE) to use a POSIX-friendly sed extraction of trailing digits (use sed with extended regex to capture the final numeric group and return only that capture) instead of grep -oP so the script works on macOS and Linux.
254-273: Add language specifiers to text output blocks.Several code blocks (lines 254, 336, 403, 467, 478, 558) lack language specifiers. For text output and prompts, use
```textto satisfy MD040.🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@plugins/docs-tools/commands/docs-cherry-pick.md` around lines 254 - 273, The markdown contains several fenced code blocks showing plain text output (e.g., the "Cherry-pick plan for <source> → <target-branch>:" block and other prompt/output blocks referenced at lines 254, 336, 403, 467, 478, 558) that lack language specifiers; update each of those triple-backtick fences to use ```text so the blocks are marked as plain text (search for the exact text snippets like "Cherry-pick plan for <source> → <target-branch>:" and the other prompt/output blocks and replace their opening ``` with ```text). Ensure every plain-output/code block in docs-cherry-pick.md uses ```text to satisfy MD040.plugins/dita-tools/README.md (1)
35-36: Consider removing this duplicate note.This Important note duplicates the warning block on lines 8-10 almost verbatim. Having the same guidance in two places creates maintenance burden—if the instruction changes, both need updating. Consider keeping only the warning block at the top, which is more prominent.
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@plugins/dita-tools/README.md` around lines 35 - 36, Remove the duplicate "**Important:** Always validate your reworked content with the AsciiDocDITA Vale style..." note (the repeated warning block in the README) and keep the original, more prominent warning at the top; ensure only the single warning block with the Vale command ("vale --config=.vale.ini --glob='*.adoc'") remains so maintenance and guidance stay centralized.plugins/docs-tools/skills/docs-branch-audit/scripts/branch_audit.sh (2)
104-105:grep -oPmay not be portable to macOS.The
-Pflag for Perl-compatible regex is not available on macOS's default BSD grep. Consider usingsedorawkfor extraction, or document GNU grep as a prerequisite.♻️ Alternative using sed
- PR_NUMBER=$(echo "$PR_URL" | grep -oP '/pull/\K[0-9]+') - OWNER_REPO=$(echo "$PR_URL" | grep -oP 'github\.com/\K[^/]+/[^/]+') + PR_NUMBER=$(echo "$PR_URL" | sed -n 's|.*/pull/\([0-9]*\).*|\1|p') + OWNER_REPO=$(echo "$PR_URL" | sed -n 's|.*github\.com/\([^/]*/[^/]*\).*|\1|p')🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@plugins/docs-tools/skills/docs-branch-audit/scripts/branch_audit.sh` around lines 104 - 105, The two extraction commands setting PR_NUMBER and OWNER_REPO use grep -oP (Perl regex) which isn't available on macOS; replace them with portable alternatives (e.g., use sed or awk) so PR_NUMBER and OWNER_REPO are extracted reliably across platforms — update the assignments for the variables PR_NUMBER and OWNER_REPO in branch_audit.sh to use a POSIX-compatible sed/awk expression that captures the digits after "/pull/" and the owner/repo after "github.com/" respectively, preserving existing behavior and error handling.
209-220: JSON output lacks proper escaping.File paths are embedded directly into JSON strings without escaping. While unlikely for
.adocpaths, any path containing quotes, backslashes, or control characters would produce malformed JSON.If robustness is needed, consider piping through
jqfor proper encoding:printf '%s' "${INCLUDE_FILES[$i]}" | jq -Rs '.'🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@plugins/docs-tools/skills/docs-branch-audit/scripts/branch_audit.sh` around lines 209 - 220, The JSON emission in branch_audit.sh builds arrays using raw values from INCLUDE_FILES and EXCLUDE_FILES which can yield invalid JSON if paths contain quotes/backslashes; update the loops that print these arrays to JSON-escape each filename (e.g., pipe each "${INCLUDE_FILES[$i]}" and "${EXCLUDE_FILES[$i]}" through jq -Rs or a small json_escape helper) and use the jq-escaped output when echoing the quoted entries so all produced JSON strings are correctly encoded.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@plugins/docs-tools/skills/docs-branch-audit/scripts/deep_audit.sh`:
- Around line 93-101: The divergence thresholds in deep_audit.sh (the
BRANCH_DIFF_LINES checks that set CONFIDENCE and append to ISSUES) do not match
SKILL.md; change the numeric cutoffs to match the documented ranges (HIGH: <20,
MEDIUM: 20–99, NEEDS-REVIEW: >=100) so the conditional branches that set
CONFIDENCE ("high"/"medium"/"needs-review") and the ISSUES messages reflect
those ranges; update the comparisons around BRANCH_DIFF_LINES and the
corresponding ISSUES text to use 20 and 100 as the boundaries (or alternatively
update SKILL.md if you intend to keep current numbers).
In `@plugins/docs-tools/skills/docs-branch-audit/SKILL.md`:
- Around line 103-107: Documentation thresholds in SKILL.md are inconsistent
with the script deep_audit.sh; update the table so the Level ranges match the
script (HIGH should cover files with up to 50 lines differing, MEDIUM should be
51–200 lines differing, and NEEDS-REVIEW should be >200 lines differing), adjust
the Meaning text accordingly (e.g., "File identical or ≤50 lines differ" for
HIGH, "51–200 lines differ, minor/conditional conflicts" for MEDIUM, ">200 lines
differ, structural changes" for NEEDS-REVIEW), and ensure the Action column
descriptions reflect the script behavior for cherry-pick/apply expectations.
---
Nitpick comments:
In `@plugins/dita-tools/README.md`:
- Around line 35-36: Remove the duplicate "**Important:** Always validate your
reworked content with the AsciiDocDITA Vale style..." note (the repeated warning
block in the README) and keep the original, more prominent warning at the top;
ensure only the single warning block with the Vale command ("vale
--config=.vale.ini --glob='*.adoc'") remains so maintenance and guidance stay
centralized.
In `@plugins/docs-tools/commands/docs-cherry-pick.md`:
- Around line 614-617: Update the two Related entries to use fully qualified
skill names in the plugin:skill format; replace "docs-branch-audit skill" with
"plugin:docs-branch-audit" and "git-pr-reader skill" with "plugin:git-pr-reader"
(preserve the descriptive text/links and formatting). Ensure the markdown list
under "## Related" uses the exact plugin:skill identifiers so they follow the
project's naming guideline.
- Around line 1-6: Add the recommended documentation sections to the command doc
by inserting top-level headings "## Name", "## Synopsis", "## Description", and
"## Implementation" after the existing YAML frontmatter in
plugins/docs-tools/commands/docs-cherry-pick.md; populate "## Name" with a short
one-line command description (e.g., docs-cherry-pick — Backport documentation
changes to enterprise branches), "## Synopsis" with example CLI usages (showing
both PR URL and --commit patterns and common flags like --target, --dry-run,
--deep, --no-push, --ticket), "## Description" with a concise paragraph
describing the command behavior (intelligently cherry-picks/backports docs
changes and excludes files not present on target releases), and "##
Implementation" with a note or the existing workflow details that explain how
the command performs the backporting so the file matches the suggested
structure.
- Line 105: The PR_NUMBER extraction uses GNU grep's -oP which is not portable;
update the assignment that sets PR_NUMBER from SOURCE (the line referencing
PR_NUMBER and SOURCE) to use a POSIX-friendly sed extraction of trailing digits
(use sed with extended regex to capture the final numeric group and return only
that capture) instead of grep -oP so the script works on macOS and Linux.
- Around line 254-273: The markdown contains several fenced code blocks showing
plain text output (e.g., the "Cherry-pick plan for <source> → <target-branch>:"
block and other prompt/output blocks referenced at lines 254, 336, 403, 467,
478, 558) that lack language specifiers; update each of those triple-backtick
fences to use ```text so the blocks are marked as plain text (search for the
exact text snippets like "Cherry-pick plan for <source> → <target-branch>:" and
the other prompt/output blocks and replace their opening ``` with ```text).
Ensure every plain-output/code block in docs-cherry-pick.md uses ```text to
satisfy MD040.
In `@plugins/docs-tools/skills/docs-branch-audit/scripts/branch_audit.sh`:
- Around line 104-105: The two extraction commands setting PR_NUMBER and
OWNER_REPO use grep -oP (Perl regex) which isn't available on macOS; replace
them with portable alternatives (e.g., use sed or awk) so PR_NUMBER and
OWNER_REPO are extracted reliably across platforms — update the assignments for
the variables PR_NUMBER and OWNER_REPO in branch_audit.sh to use a
POSIX-compatible sed/awk expression that captures the digits after "/pull/" and
the owner/repo after "github.com/" respectively, preserving existing behavior
and error handling.
- Around line 209-220: The JSON emission in branch_audit.sh builds arrays using
raw values from INCLUDE_FILES and EXCLUDE_FILES which can yield invalid JSON if
paths contain quotes/backslashes; update the loops that print these arrays to
JSON-escape each filename (e.g., pipe each "${INCLUDE_FILES[$i]}" and
"${EXCLUDE_FILES[$i]}" through jq -Rs or a small json_escape helper) and use the
jq-escaped output when echoing the quoted entries so all produced JSON strings
are correctly encoded.
In `@plugins/docs-tools/skills/docs-branch-audit/scripts/deep_audit.sh`:
- Around line 141-157: The current git apply --check --3way invocation (when
checking "$FILE_PATCH") operates against the current worktree, not the target
branch snapshot; update the deep_audit.sh logic around the FILE_PATCH check to
either (A) add a clear comment documenting this limitation as a best-effort
heuristic, or (B) implement a safer check by building an explicit temporary
index/tree from the target ref and running git apply --check --3way against that
index (use the same FILE_PATCH and preserve setting of CONFIDENCE and ISSUES),
and ensure the CONFIDENCE and ISSUES handling (patch-conflict / patch-mismatch /
patch: applies cleanly) remains unchanged.
In `@plugins/docs-tools/skills/docs-branch-audit/SKILL.md`:
- Around line 61-75: The fenced code blocks showing the branch audit output
(e.g., the block starting with "=== Branch Audit: enterprise-4.17 ===" and other
similar audit/example blocks) are missing a language specifier; update each
triple-backtick fence to include a language such as text or console (for example
replace ``` with ```text) so the linter MD040 is satisfied for the blocks
referenced around the "Branch Audit" examples.
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
Run ID: 2222bbbd-c4c5-458b-bfc1-b9f14596199c
📒 Files selected for processing (5)
plugins/dita-tools/README.mdplugins/docs-tools/commands/docs-cherry-pick.mdplugins/docs-tools/skills/docs-branch-audit/SKILL.mdplugins/docs-tools/skills/docs-branch-audit/scripts/branch_audit.shplugins/docs-tools/skills/docs-branch-audit/scripts/deep_audit.sh
plugins/docs-tools/skills/docs-branch-audit/scripts/deep_audit.sh
Outdated
Show resolved
Hide resolved
There was a problem hiding this comment.
Actionable comments posted: 3
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@plugins/docs-tools/commands/docs-cherry-pick.md`:
- Line 296: Add the missing language identifiers to unlabeled fenced code blocks
in plugins/docs-tools/commands/docs-cherry-pick.md by changing each
triple-backtick fence that currently has no info string to use "text" (e.g.,
```text), specifically for the blocks containing the "Cherry-pick plan for
<source> → <target-branch>:" section, the Agent(...) example
(Agent(subagent_type="general-purpose", description="resolve conflict in
<file>", ...)), the "The following conflicts could not be resolved
automatically:" block, the "Stats comparison (source PR vs backport):" block and
the "⚠️ Stats divergence..." / "Branch pushed: <branch-name>" blocks; apply the
same fix at the additional locations mentioned (lines 378, 445, 509, 520, 600)
so markdownlint MD040 is satisfied.
- Line 74: Replace bare inline skill references with fully qualified
plugin:skill names; specifically change occurrences like `docs-branch-audit` to
the fully qualified form (for example `docs-tools:docs-branch-audit`) and update
the other mentioned inline references similarly (e.g., replace any bare
`jira-reader` style mentions with `docs-tools:jira-reader` or the appropriate
plugin prefix). Ensure all agent frontmatter, Skill tool invocation examples,
inline text references, and cross-references in this document use the
plugin:skill syntax consistently.
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
Run ID: 16f08627-a69e-45ef-aa36-45a1999f8b86
📒 Files selected for processing (1)
plugins/docs-tools/commands/docs-cherry-pick.md
There was a problem hiding this comment.
♻️ Duplicate comments (1)
plugins/docs-tools/commands/docs-cherry-pick.md (1)
302-322:⚠️ Potential issue | 🟡 MinorAdd language identifiers to unlabeled fenced code blocks.
These fences still violate MD040 (
fenced-code-language). Please label them (e.g.,text) to keep lint clean.Suggested fix
-``` +```text Cherry-pick plan for <source> → <target-branch>: ... -``` +``` -``` +```text Agent(subagent_type="general-purpose", description="resolve conflict in <file>", ... -``` +``` -``` +```text The following conflicts could not be resolved automatically: ... -``` +``` -``` +```text Stats comparison (source PR vs backport): ... -``` +``` -``` +```text ⚠️ Stats divergence exceeds excluded files — review the backport for unintended changes. -``` +``` -``` +```text Branch pushed: <branch-name> ... -``` +```Also applies to: 384-418, 451-460, 515-522, 526-528, 606-615
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@plugins/docs-tools/commands/docs-cherry-pick.md` around lines 302 - 322, The markdown file contains multiple unlabeled fenced code blocks (e.g., the block starting with "Cherry-pick plan for <source> → <target-branch>:", the block containing 'Agent(subagent_type="general-purpose"...', the block beginning "The following conflicts could not be resolved automatically:", the "Stats comparison (source PR vs backport):" block, and the "Branch pushed: <branch-name>" block) which trigger MD040; update each triple-backtick fence to include a language identifier (suggest `text`) so they read ```text ... ```; apply the same fix to the other similar unlabeled blocks mentioned in the comment so all code fences in docs-cherry-pick.md are labeled.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Duplicate comments:
In `@plugins/docs-tools/commands/docs-cherry-pick.md`:
- Around line 302-322: The markdown file contains multiple unlabeled fenced code
blocks (e.g., the block starting with "Cherry-pick plan for <source> →
<target-branch>:", the block containing
'Agent(subagent_type="general-purpose"...', the block beginning "The following
conflicts could not be resolved automatically:", the "Stats comparison (source
PR vs backport):" block, and the "Branch pushed: <branch-name>" block) which
trigger MD040; update each triple-backtick fence to include a language
identifier (suggest `text`) so they read ```text ... ```; apply the same fix to
the other similar unlabeled blocks mentioned in the comment so all code fences
in docs-cherry-pick.md are labeled.
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
Run ID: ac9ea762-4ca1-4fdb-8014-e5c935c68151
📒 Files selected for processing (1)
plugins/docs-tools/commands/docs-cherry-pick.md
Member
|
This seems to be an openshift-docs specific workflow. Has it been tested against other docs repos? |
Author
|
tested against OCP - would love if tested against other repos could this be refined over time by others or are we creating a doc specific repo |
Member
|
docs-branch-audit looks great. docs-cherry-pick needs some work I think. The cherry-pick logic relies on Claude interpreting 631 lines of prose, not ideal. Also, Claude commands are deprecated, https://code.claude.com/docs/en/skills We need to migrate away from these generally (plans underway to get rid of the commands in this repo). |
Author
|
Audit one is a nice heads up for all this vale stuff so I need to work CP stuff command as we are moving away from this. CP was use when lots of changes |
Author
|
@aireilly updated based on your feedback found --dryrun useful for large PRs with lots of changes. There is a sweet spot:
|
kquinn1204
force-pushed
the
add-docs-cherry-pick-command
branch
from
1a303e1 to
e3b7ef5
Compare
- Update install links to marketplace URL across all plugin READMEs - Add AsciiDocDITA Vale validation note to dita-tools README and both rework modes in dita-rework command Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> rh-pre-commit.version: 2.3.2 rh-pre-commit.check-secrets: ENABLED
Add intelligent cherry-pick backporting for documentation changes across enterprise branches. The command audits file existence and content compatibility per branch, then creates targeted backport PRs that exclude modules not present on each release. Two components: 1. docs-branch-audit skill - Checks which files from a PR/commit exist on target branches, with optional deep content comparison that reports confidence levels (HIGH/MEDIUM/NEEDS-REVIEW) per file. 2. docs-cherry-pick command - Full backport workflow: audit, cherry-pick, version-aware conflict resolution, stats comparison, and PR description generation. Supports --dry-run --deep for audit-only mode. The GitHub cherry-pick bot handles simple cases (all files exist, no conflicts). This command fills the gap for PRs with file exclusions, path differences between releases, or version-specific content that requires context-aware conflict resolution. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
HIGH: < 20 lines, MEDIUM: 20-99 lines, NEEDS-REVIEW: 100+ lines Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Restructure docs-cherry-pick.md to match the command format expected by the plugin linter (.claudelint.yaml). Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add GitLab merge-requests refspec for open MR mode alongside the existing GitHub pull refspec. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Fixes markdownlint MD040 warnings. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Move cherry-pick logic from 631 lines of prose (command) into a bash script that handles all deterministic operations: PR info fetching, branch audit, cherry-pick execution, commit, push, and PR description generation. The SKILL.md is a thin wrapper (123 lines) covering only conflict resolution rules that require LLM judgement. - Delete commands/docs-cherry-pick.md - Add skills/docs-cherry-pick/SKILL.md - Add skills/docs-cherry-pick/scripts/cherry_pick.sh Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> and counted toward included files rather than excluded. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
kquinn1204
force-pushed
the
add-docs-cherry-pick-command
branch
from
e3b7ef5 to
ea87401
Compare
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Adds
docs-cherry-pickcommand anddocs-branch-auditskill for intelligent cherry-pick backporting across enterprise branches.What it does
--deepmode scores each file as HIGH/MEDIUM/NEEDS-REVIEW based on content divergenceWhen to use this vs the GitHub bot
Use the bot for simple backports (no exclusions, no path changes). Use this command when:
Tested against 7 real PRs — 6 were bot-suitable, 1 (PR #106280, 28 PTP files) required this command due to exclusions, path differences, and 440-line content divergence.
Files added
plugins/docs-tools/commands/docs-cherry-pick.md— command orchestrationplugins/docs-tools/skills/docs-branch-audit/SKILL.md— skill definitionplugins/docs-tools/skills/docs-branch-audit/scripts/branch_audit.sh— file existence checkerplugins/docs-tools/skills/docs-branch-audit/scripts/deep_audit.sh— content divergence analyzerDepends on the existing
git-pr-readerskill. Requiresgit,python3,bash.Summary by CodeRabbit