@ericrisco/rsc 1.1.2 → 1.1.3
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/README.md +6 -6
- package/manifest.json +1 -1
- package/package.json +1 -1
- package/scripts/lib/eval-integrity.js +1 -1
package/README.md
CHANGED
|
@@ -6,7 +6,7 @@
|
|
|
6
6
|
|
|
7
7
|
[](https://www.npmjs.com/package/@ericrisco/rsc)
|
|
8
8
|
[](https://www.npmjs.com/package/@ericrisco/rsc)
|
|
9
|
-
[](#the-catalog)
|
|
10
10
|
[](LICENSE)
|
|
11
11
|
[](https://github.com/ericrisco/rsc-harness/stargazers)
|
|
12
12
|
|
|
@@ -17,7 +17,7 @@ skills that fit — one at a time — into every assistant you pick, and keeps t
|
|
|
17
17
|
equipped as you work.
|
|
18
18
|
|
|
19
19
|
From *"document my company"* to *"ship a FastAPI service"* to *"grow my YouTube
|
|
20
|
-
channel"* — **
|
|
20
|
+
channel"* — **281 skills across 23 domains**, every one researched against live
|
|
21
21
|
2025-2026 sources and **adversarially scored ≥ 8.5/10** before it shipped.
|
|
22
22
|
|
|
23
23
|
```bash
|
|
@@ -26,7 +26,7 @@ npx @ericrisco/rsc # plain-language wizard — no jargon, installs wh
|
|
|
26
26
|
|
|
27
27
|
<br>
|
|
28
28
|
|
|
29
|
-
<img src="https://raw.githubusercontent.com/ericrisco/rsc-harness/main/site/meta-harness.png" alt="rsc — the model (Claude, GPT, Gemini) is the brain your coding agent rents, and you already have that. Claude Code, Codex and opencode sit in the middle. Below them the rsc meta-harness adds everything the model lacks: memory (02-DOCS/) so it stops inventing, arms (01-TOOLS/) so it can execute, a trade (
|
|
29
|
+
<img src="https://raw.githubusercontent.com/ericrisco/rsc-harness/main/site/meta-harness.png" alt="rsc — the model (Claude, GPT, Gemini) is the brain your coding agent rents, and you already have that. Claude Code, Codex and opencode sit in the middle. Below them the rsc meta-harness adds everything the model lacks: memory (02-DOCS/) so it stops inventing, arms (01-TOOLS/) so it can execute, a trade (281 skills) so it stops guessing, and reflexes (deterministic hooks) so it cannot forget the rule mid-session." width="960">
|
|
30
30
|
|
|
31
31
|
**The model is the only part your agent already has.** `rsc` is the meta-harness that adds
|
|
32
32
|
the rest: a **memory** so it stops inventing, **arms** so it can execute, a **trade** so it stops
|
|
@@ -139,7 +139,7 @@ $ rsc
|
|
|
139
139
|
██████╗ ███████╗ ██████╗ ← animated gradient wordmark
|
|
140
140
|
██╔══██╗██╔════╝██╔════╝
|
|
141
141
|
██████╔╝███████╗██║
|
|
142
|
-
|
|
142
|
+
281 skills · one CLI · zero bloat
|
|
143
143
|
|
|
144
144
|
What do you want to do? ↑↓ move · enter select
|
|
145
145
|
❯ Base install — the essentials (8 skills)
|
|
@@ -187,7 +187,7 @@ rsc add youtube-api remotion-video # …grow a channel, edit with Remotion
|
|
|
187
187
|
rsc add fastapi --target claude,codex # install into several assistants at once
|
|
188
188
|
rsc install --profile minimal # the base: orient + suggest + bro + unslop + show-me + eli5 + harness + init
|
|
189
189
|
rsc install --profile core # floor + the full SDD workflow
|
|
190
|
-
rsc install --profile full # everything (all
|
|
190
|
+
rsc install --profile full # everything (all 281 skills)
|
|
191
191
|
rsc install --profile full --without go
|
|
192
192
|
rsc consult "I want to launch a SaaS" # recommend only, no install
|
|
193
193
|
rsc registry refresh # write .rsc/skill-registry.{json,md}
|
|
@@ -253,7 +253,7 @@ just asks in plain language.
|
|
|
253
253
|
|
|
254
254
|
## The catalog
|
|
255
255
|
|
|
256
|
-
|
|
256
|
+
281 skills, grouped by what you're trying to do. Click any skill to read its
|
|
257
257
|
`SKILL.md`. It fires on its own when a task matches.
|
|
258
258
|
|
|
259
259
|
### 🧭 Core & control plane
|
package/manifest.json
CHANGED
package/package.json
CHANGED
|
@@ -9,7 +9,7 @@
|
|
|
9
9
|
//
|
|
10
10
|
// A lift measured against a control that read the answer key is a lower bound of unknown size.
|
|
11
11
|
// Publishing it as a number is the claim-without-evidence this repo keeps paying for (P2), and it
|
|
12
|
-
// feeds skill-harden's --holdout, which decides what enters a
|
|
12
|
+
// feeds skill-harden's --holdout, which decides what enters a 281-skill catalog.
|
|
13
13
|
//
|
|
14
14
|
// This module answers the question deterministically, from the transcripts themselves — never from
|
|
15
15
|
// what an agent says about its own behaviour, which is the way this fails open.
|