replen 1.0.37 → 1.0.38
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/extras/skills/replen/SKILL.md +26 -10
- package/package.json +1 -1
|
@@ -187,23 +187,39 @@ mcp__replen__replen_capture_insight(
|
|
|
187
187
|
)
|
|
188
188
|
```
|
|
189
189
|
|
|
190
|
-
### Step 4 — Present
|
|
190
|
+
### Step 4 — Present ONLY the wins (record everything, surface what works)
|
|
191
191
|
|
|
192
|
-
|
|
192
|
+
You already RECORDED every verdict in Step 3d, including skips — those are
|
|
193
|
+
load-bearing (relevance-floor calibration, repo_quality, modality suppression,
|
|
194
|
+
the Activity feed). But **do NOT narrate the skips to the user.** A wall of
|
|
195
|
+
"skip, skip, skip, but…" makes Replen look like it's grasping. Present like an
|
|
196
|
+
Apple keynote: lead with what's good, stay quiet about the rest.
|
|
197
|
+
|
|
198
|
+
**Surface ONLY the wins** — `adopt` / `port` / `cherry-pick` / `clean-room` /
|
|
199
|
+
`upgrade` + any `lesson` / `boundary`. Skips and defers are recorded silently and
|
|
200
|
+
never listed. Lead with the strongest (an `upgrade` to something they already
|
|
201
|
+
have usually beats a new `adopt`).
|
|
202
|
+
|
|
203
|
+
If there are wins:
|
|
193
204
|
|
|
194
205
|
```
|
|
195
|
-
|
|
196
|
-
|
|
197
|
-
|
|
198
|
-
|
|
199
|
-
✗ skip: vercel/turbo (deep) — wrong runtime, already on Vite
|
|
206
|
+
For tech-news-site:
|
|
207
|
+
⬆ upgrade: someorg/fast-og — beats your og-image render (streams + caches vs your sync redraw) · moderate
|
|
208
|
+
✓ adopt: kribblo/node-ffmpeg-installer — drop-in ffmpeg-static swap · quick
|
|
209
|
+
💡 lesson: borrow graph-as-memory premise, link repos not files (→ Atlas) [via graphify/graphify]
|
|
200
210
|
|
|
201
|
-
|
|
202
|
-
— captured to Atlas; shown only when a Pass-3/4 insight actually fired
|
|
211
|
+
(Triaged 9 candidates; surfacing the 2 worth acting on + 1 idea worth keeping.)
|
|
203
212
|
|
|
204
|
-
|
|
213
|
+
What would you like to do with each? (star / hide / handoff)
|
|
205
214
|
```
|
|
206
215
|
|
|
216
|
+
The one-line "(Triaged N…)" footer is the *only* acknowledgement that skips
|
|
217
|
+
happened — honest, but it doesn't parade them.
|
|
218
|
+
|
|
219
|
+
**If there are NO wins, say nothing** — or at most one calm line ("Nothing
|
|
220
|
+
actionable for `<repo>` today — N triaged."). Never list the skips. Silence on a
|
|
221
|
+
quiet day IS the calm-cadence contract, not a failure.
|
|
222
|
+
|
|
207
223
|
Then, for each candidate, capture the user's choice. For each action,
|
|
208
224
|
POST to `/api/state`:
|
|
209
225
|
|
package/package.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
{
|
|
2
2
|
"name": "replen",
|
|
3
|
-
"version": "1.0.
|
|
3
|
+
"version": "1.0.38",
|
|
4
4
|
"description": "Make your AI coding tools smarter. One command, no API keys, free. Replen watches what your projects actually do and surfaces a few things worth bringing in each month. Use one as is, port a piece of another, cherry pick an idea, or build it clean room. The match happens inside your AI tool's session. A few actionable matches a month, by design.",
|
|
5
5
|
"type": "module",
|
|
6
6
|
"bin": {
|