refresh-index

Name: refresh-index
Author: luongnv89

742stars61forks

How to install refresh-index?

luongnv89/asm/refresh-index

$npx -y skills add luongnv89/asm --skill refresh-index

Installs into the current project.

Summary

Re-ingest every already-enabled repo in data/skill-index-resources.json: sync the working tree, run preindex, rebuild the catalog for verification, summarize updated/unchanged/failed/skipped, then gate commit + PR behind explicit confirmation. Use when asked to refresh the index,

Files · 1

View on GitHub

SKILL.md

1	# Refresh Index
2
3	You are refreshing every enabled repository in the ASM curated skill index. The goal is to re-ingest each source so `data/skill-index/{owner}_{repo}.json` reflects the latest upstream state, classify each repo as updated / unchanged / failed / skipped, verify the catalog still rebuilds cleanly, and open a single PR that ships only the data changes — with explicit user confirmation before any commit or push.
4
5	This is the inverse of `skill-index-updater`. That skill adds new repos from URLs the user supplies. This skill refreshes the repos that are already enabled in `data/skill-index-resources.json`.
6
7	## Repo Sync Before Edits (mandatory)
8
9	Before re-ingesting anything, pull the latest remote branch:
10
11	```bash
12	branch="$(git rev-parse --abbrev-ref HEAD)"
13	dirty=0
14	if [ -n "$(git status --porcelain)" ]; then
15	git stash push -u -m "pre-refresh-index: ${branch}"
16	dirty=1
17	fi
18	git fetch origin
19	git pull --rebase origin "$branch"
20	if [ "$dirty" -eq 1 ]; then
21	git stash pop \|\| {
22	echo "✗ Stash pop failed — recover with: git stash list && git stash show -p stash@{0}"
23	exit 1
24	}
25	fi
26	```
27
28	If `origin` is missing or rebase conflicts occur, stop and ask the user before continuing. Never silently overwrite local edits to `data/skill-index/` or `data/skill-index-resources.json`.
29
30	## Prerequisites
31
32	Verify each before any ingest. Stop and tell the user if any fails.
33
34	- `node` and `npm` on PATH (`command -v node`, `command -v npm`) — required by `npm run preindex`
35	- `asm` on PATH (`command -v asm`) — invoked transitively by the ingester for `asm eval`
36	- `gh` on PATH and authenticated (`gh auth status`) — required for PR creation
37	- `git` on PATH and inside the ASM repo working tree (`git rev-parse --show-toplevel`)
38	- Network access to `github.com` (each ingest clones the upstream repo)
39
40	## Pipeline
41
42	Follow these steps in order. Each step has a verification check — do not proceed if the check fails.
43
44	### Step 1: Enumerate the index
45
46	Read `data/skill-index-resources.json` and split entries into two lists:
47
48	```bash
49	ROOT="$(git rev-parse --show-toplevel)"
50	RES="$ROOT/data/skill-index-resources.json"
51	```
52
53	Use `jq` (or your equivalent JSON reader) to build two lists, keyed on the `source` string (`github:owner/repo`) because that is the identifier `preindex` echoes in its log lines:
54
55	- `enabled[]` — every repo with `"enabled": true`. These will be refreshed.
56	- `disabled[]` — every repo with `"enabled": false`. These land in the skipped bucket up front with reason `"disabled in skill-index-resources.json"`.
57
58	```bash
59	jq -r '.repos[] \| select(.enabled == true) \| .source' "$RES" # e.g. github:anthropics/skills
60	jq -r '.repos[] \| select(.enabled == false) \| .source' "$RES"
61	```
62
63	Also derive the on-disk file key (`{owner}_{repo}`) for each enabled repo — this is what `data/skill-index/{owner}_{repo}.json` is named:
64
65	```bash
66	jq -r '.repos[] \| select(.enabled == true) \| "\(.source)\t\(.owner)_\(.repo)"' "$RES"
67	```
68
69	Verification: both lists are non-empty (the index always contains at least one enabled repo and one disabled self-reference). Also validate the existing per-repo JSON is parseable before re-ingesting — a pre-existing corruption otherwise gets masked by the refresh:
70
71	```bash
72	jq empty data/skill-index/*.json # exits non-zero if any file is invalid
73	```
74
75	If `enabled[]` is empty or pre-validation fails, stop — there is nothing safe to refresh.
76
77	### Step 2: Snapshot pre-run skill counts
78
79	Before re-ingesting, record the current `skillCount` for each enabled repo. This lets Step 7 report skill-count deltas.
80
81	```bash
82	mkdir -p /tmp/refresh-index
83	SNAPSHOT="/tmp/refresh-index/pre-snapshot.json"
84	echo "{}" > "$SNAPSHOT"
85	for entry in $(jq -r '.repos[] \| select(.enabled == true) \| "\(.owner)_\(.repo)"' "$RES"); do
86	file="$ROOT/data/skill-index/${entry}.json"
87	if [ -f "$file" ]; then
88	count=$(jq -r '.skillCount // 0' "$file")
89	jq --arg k "$entry" --argjson v "$count" '. + {($k): $v}' "$SNAPSHOT" > "$SNAPSHOT.tmp" && mv "$SNAPSHOT.tmp" "$SNAPSHOT"
90	fi
91	done
92	```
93
94	If a repo is enabled but has no existing index file, treat the pre-count as `0`. The repo will then show up in updated with a positive delta in Step 7.
95
96	### Step 3: Run `npm run preindex`
97
98	Re-ingest every enabled

People who liked this also viewed

6 picks

Type

refresh-index

742stars61forks

View on GitHub

How to install refresh-index?

luongnv89/asm/refresh-index

$npx -y skills add luongnv89/asm --skill refresh-index

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Refresh Index
2
3	You are refreshing every enabled repository in the ASM curated skill index. The goal is to re-ingest each source so `data/skill-index/{owner}_{repo}.json` reflects the latest upstream state, classify each repo as updated / unchanged / failed / skipped, verify the catalog still rebuilds cleanly, and open a single PR that ships only the data changes — with explicit user confirmation before any commit or push.
4
5	This is the inverse of `skill-index-updater`. That skill adds new repos from URLs the user supplies. This skill refreshes the repos that are already enabled in `data/skill-index-resources.json`.
6
7	## Repo Sync Before Edits (mandatory)
8
9	Before re-ingesting anything, pull the latest remote branch:
10
11	```bash
12	branch="$(git rev-parse --abbrev-ref HEAD)"
13	dirty=0
14	if [ -n "$(git status --porcelain)" ]; then
15	git stash push -u -m "pre-refresh-index: ${branch}"
16	dirty=1
17	fi
18	git fetch origin
19	git pull --rebase origin "$branch"
20	if [ "$dirty" -eq 1 ]; then
21	git stash pop \|\| {
22	echo "✗ Stash pop failed — recover with: git stash list && git stash show -p stash@{0}"
23	exit 1
24	}
25	fi
26	```
27
28	If `origin` is missing or rebase conflicts occur, stop and ask the user before continuing. Never silently overwrite local edits to `data/skill-index/` or `data/skill-index-resources.json`.
29
30	## Prerequisites
31
32	Verify each before any ingest. Stop and tell the user if any fails.
33
34	- `node` and `npm` on PATH (`command -v node`, `command -v npm`) — required by `npm run preindex`
35	- `asm` on PATH (`command -v asm`) — invoked transitively by the ingester for `asm eval`
36	- `gh` on PATH and authenticated (`gh auth status`) — required for PR creation
37	- `git` on PATH and inside the ASM repo working tree (`git rev-parse --show-toplevel`)
38	- Network access to `github.com` (each ingest clones the upstream repo)
39
40	## Pipeline
41
42	Follow these steps in order. Each step has a verification check — do not proceed if the check fails.
43
44	### Step 1: Enumerate the index
45
46	Read `data/skill-index-resources.json` and split entries into two lists:
47
48	```bash
49	ROOT="$(git rev-parse --show-toplevel)"
50	RES="$ROOT/data/skill-index-resources.json"
51	```
52
53	Use `jq` (or your equivalent JSON reader) to build two lists, keyed on the `source` string (`github:owner/repo`) because that is the identifier `preindex` echoes in its log lines:
54
55	- `enabled[]` — every repo with `"enabled": true`. These will be refreshed.
56	- `disabled[]` — every repo with `"enabled": false`. These land in the skipped bucket up front with reason `"disabled in skill-index-resources.json"`.
57
58	```bash
59	jq -r '.repos[] \| select(.enabled == true) \| .source' "$RES" # e.g. github:anthropics/skills
60	jq -r '.repos[] \| select(.enabled == false) \| .source' "$RES"
61	```
62
63	Also derive the on-disk file key (`{owner}_{repo}`) for each enabled repo — this is what `data/skill-index/{owner}_{repo}.json` is named:
64
65	```bash
66	jq -r '.repos[] \| select(.enabled == true) \| "\(.source)\t\(.owner)_\(.repo)"' "$RES"
67	```
68
69	Verification: both lists are non-empty (the index always contains at least one enabled repo and one disabled self-reference). Also validate the existing per-repo JSON is parseable before re-ingesting — a pre-existing corruption otherwise gets masked by the refresh:
70
71	```bash
72	jq empty data/skill-index/*.json # exits non-zero if any file is invalid
73	```
74
75	If `enabled[]` is empty or pre-validation fails, stop — there is nothing safe to refresh.
76
77	### Step 2: Snapshot pre-run skill counts
78
79	Before re-ingesting, record the current `skillCount` for each enabled repo. This lets Step 7 report skill-count deltas.
80
81	```bash
82	mkdir -p /tmp/refresh-index
83	SNAPSHOT="/tmp/refresh-index/pre-snapshot.json"
84	echo "{}" > "$SNAPSHOT"
85	for entry in $(jq -r '.repos[] \| select(.enabled == true) \| "\(.owner)_\(.repo)"' "$RES"); do
86	file="$ROOT/data/skill-index/${entry}.json"
87	if [ -f "$file" ]; then
88	count=$(jq -r '.skillCount // 0' "$file")
89	jq --arg k "$entry" --argjson v "$count" '. + {($k): $v}' "$SNAPSHOT" > "$SNAPSHOT.tmp" && mv "$SNAPSHOT.tmp" "$SNAPSHOT"
90	fi
91	done
92	```
93
94	If a repo is enabled but has no existing index file, treat the pre-count as `0`. The repo will then show up in updated with a positive delta in Step 7.
95
96	### Step 3: Run `npm run preindex`
97
98	Re-ingest every enabled

People who liked this also viewed

6 picks

Type