github-sync-upstream

Name: github-sync-upstream
Author: opendatahub-io

36stars74forks

How to install github-sync-upstream?

opendatahub-io/ai-helpers/github-sync-upstream

$npx -y skills add opendatahub-io/ai-helpers --skill github-sync-upstream

Installs into the current project.

Summary

Sync code from an upstream GitHub repository into a target fork (e.g., opendatahub-io midstream). Detects remotes from the current repo, or clones fresh if run from outside. Fetches upstream, merges into a sync branch, restores protected files, resolves conflicts, and opens a PR

Files · 1

View on GitHub

SKILL.md

1	# Sync Upstream
2
3	Merge upstream commits into a sync branch on the user's fork and open a
4	PR to the target repo. See `references/workflow.md` for exact script
5	invocations, exit-code handling, and the summary template.
6
7	Commit SHA (optional): `$ARGUMENTS`
8
9	## Step 1: Determine Working Context
10
11	```bash
12	REPO_ROOT=$(git rev-parse --show-toplevel 2>/dev/null) \|\| REPO_ROOT=""
13	```
14
15	If `REPO_ROOT` is non-empty, run `scripts/detect-remotes.sh --repo "${REPO_ROOT}"`
16	and ask via `AskUserQuestion` whether this is the correct repo.
17	If yes → Step 2A. If no or not in a repo → Step 2B.
18
19	## Step 2A: In-Repo Setup
20
21	Pre-fill upstream/target from detect-remotes output. Ask via
22	`AskUserQuestion` to confirm: upstream repo, target repo, branches
23	(default `main`). Run `scripts/setup-remotes.sh`, parse `UPSTREAM_REMOTE` and
24	`TARGET_REMOTE`. Save `ORIGINAL_BRANCH` from current HEAD.
25
26	## Step 2B: Clone From Scratch
27
28	Ask for upstream repo, target repo, and branches. If the user has a
29	local clone, use its path and continue as Step 2A. If not, ask if they
30	have a GitHub fork — run `scripts/clone-fork.sh` then `scripts/setup-remotes.sh`. If
31	no fork exists, ask them to create one and re-run. Stop.
32
33	Save `ORIGINAL_BRANCH` from current HEAD.
34
35	## Step 3: Protected Files
36
37	Protected files keep the target version, discarding upstream changes.
38	Ask via `AskUserQuestion` for glob patterns to protect (or none).
39	Suggest common examples: `OWNERS`, `.tekton/.yaml`,
40	`Dockerfile*konflux`.
41
42	## Step 4: Pre-flight Check
43
44	Verify `origin` does not point to upstream or target. If it does, tell
45	the user to set origin to their personal fork and stop.
46
47	## Step 5: Merge
48
49	Run `scripts/sync-merge.sh`. Handle exit codes 0 (success), 1 (conflicts), and
50	3 (duplicate branch) as described in `references/workflow.md`.
51
52	## Step 6: Push and Open PR
53
54	Show PR summary and ask via `AskUserQuestion`: open a PR or just push?
55	If confirmed, run `scripts/open-pr.sh`.
56
57	## Step 7: Cleanup and Summary
58
59	Check out `ORIGINAL_BRANCH`. For Path B clones, inform the user of the
60	temp directory. Display the summary per `references/workflow.md`.

People who liked this also viewed

6 picks

Type

github-sync-upstream

36stars74forks

View on GitHub

How to install github-sync-upstream?

opendatahub-io/ai-helpers/github-sync-upstream

$npx -y skills add opendatahub-io/ai-helpers --skill github-sync-upstream

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Sync Upstream
2
3	Merge upstream commits into a sync branch on the user's fork and open a
4	PR to the target repo. See `references/workflow.md` for exact script
5	invocations, exit-code handling, and the summary template.
6
7	Commit SHA (optional): `$ARGUMENTS`
8
9	## Step 1: Determine Working Context
10
11	```bash
12	REPO_ROOT=$(git rev-parse --show-toplevel 2>/dev/null) \|\| REPO_ROOT=""
13	```
14
15	If `REPO_ROOT` is non-empty, run `scripts/detect-remotes.sh --repo "${REPO_ROOT}"`
16	and ask via `AskUserQuestion` whether this is the correct repo.
17	If yes → Step 2A. If no or not in a repo → Step 2B.
18
19	## Step 2A: In-Repo Setup
20
21	Pre-fill upstream/target from detect-remotes output. Ask via
22	`AskUserQuestion` to confirm: upstream repo, target repo, branches
23	(default `main`). Run `scripts/setup-remotes.sh`, parse `UPSTREAM_REMOTE` and
24	`TARGET_REMOTE`. Save `ORIGINAL_BRANCH` from current HEAD.
25
26	## Step 2B: Clone From Scratch
27
28	Ask for upstream repo, target repo, and branches. If the user has a
29	local clone, use its path and continue as Step 2A. If not, ask if they
30	have a GitHub fork — run `scripts/clone-fork.sh` then `scripts/setup-remotes.sh`. If
31	no fork exists, ask them to create one and re-run. Stop.
32
33	Save `ORIGINAL_BRANCH` from current HEAD.
34
35	## Step 3: Protected Files
36
37	Protected files keep the target version, discarding upstream changes.
38	Ask via `AskUserQuestion` for glob patterns to protect (or none).
39	Suggest common examples: `OWNERS`, `.tekton/.yaml`,
40	`Dockerfile*konflux`.
41
42	## Step 4: Pre-flight Check
43
44	Verify `origin` does not point to upstream or target. If it does, tell
45	the user to set origin to their personal fork and stop.
46
47	## Step 5: Merge
48
49	Run `scripts/sync-merge.sh`. Handle exit codes 0 (success), 1 (conflicts), and
50	3 (duplicate branch) as described in `references/workflow.md`.
51
52	## Step 6: Push and Open PR
53
54	Show PR summary and ask via `AskUserQuestion`: open a PR or just push?
55	If confirmed, run `scripts/open-pr.sh`.
56
57	## Step 7: Cleanup and Summary
58
59	Check out `ORIGINAL_BRANCH`. For Path B clones, inform the user of the
60	temp directory. Display the summary per `references/workflow.md`.

People who liked this also viewed

6 picks

Type