company-skill 4.1.0 → 4.3.0
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 +36 -6
- package/package.json +1 -1
- package/skill/SKILL.md +17 -2
package/README.md
CHANGED
|
@@ -2,6 +2,8 @@
|
|
|
2
2
|
|
|
3
3
|
[](https://www.npmjs.com/package/company-skill) [](LICENSE) [](https://www.npmjs.com/package/company-skill)
|
|
4
4
|
|
|
5
|
+
> *You don't prompt agents one at a time. You write a team in markdown, hand them a goal, and go to sleep. In the morning, STATUS.md tells you what got done, what got rejected, and what the company learned. The playbook from session 3 makes session 4 faster. By session 10, the company runs itself better than you could direct it manually.*
|
|
6
|
+
|
|
5
7
|
**Define your team in markdown. Give it a goal. Walk away.**
|
|
6
8
|
|
|
7
9
|
A Claude Code skill that runs your entire company — CEO delegates, departments execute in parallel, built-in reviewers verify — and doesn't stop until the goal is done.
|
|
@@ -143,13 +145,20 @@ Override per employee: `- ML Scientist, experiments [opus]`
|
|
|
143
145
|
|
|
144
146
|
## Installed Skills
|
|
145
147
|
|
|
146
|
-
Auto-installed on first run.
|
|
148
|
+
Auto-installed on first run. Leads assign skills to workers by task type.
|
|
149
|
+
|
|
150
|
+
| Task type | Skill | Pack |
|
|
151
|
+
|-----------|-------|------|
|
|
152
|
+
| Code review | /review | gstack |
|
|
153
|
+
| Bug fix | /investigate | gstack |
|
|
154
|
+
| QA testing | /qa | gstack |
|
|
155
|
+
| Ship code | /ship | gstack |
|
|
156
|
+
| Browse/test site | /browse | gstack |
|
|
157
|
+
| Security audit | /secure-phase | trailofbits |
|
|
158
|
+
| Debug with state | /gsd-debug | GSD |
|
|
159
|
+
| Plan work | /gsd-plan-phase | GSD |
|
|
147
160
|
|
|
148
|
-
|
|
149
|
-
|------|-------------------|
|
|
150
|
-
| gstack | /review, /ship, /qa, /investigate, /browse |
|
|
151
|
-
| GSD | /gsd-plan-phase, /gsd-execute-phase, /gsd-verify-work, /gsd-debug |
|
|
152
|
-
| trailofbits | Security audit, vulnerability detection |
|
|
161
|
+
If no skill matches the task, workers use raw tools.
|
|
153
162
|
|
|
154
163
|
<details>
|
|
155
164
|
<summary>Install more skill packs</summary>
|
|
@@ -175,6 +184,27 @@ Auto-installed on first run. When installed, employees MUST use them.
|
|
|
175
184
|
{dept}/ Per-employee findings (persist across sessions)
|
|
176
185
|
```
|
|
177
186
|
|
|
187
|
+
## Design Choices
|
|
188
|
+
|
|
189
|
+
Three principles behind the skill:
|
|
190
|
+
|
|
191
|
+
- **One file to define the team.** COMPANY.md is the only thing you write. Everything else — delegation, task routing, quality checks — is automatic.
|
|
192
|
+
- **No iteration limit.** The loop runs until criteria.json says done. Not 3 cycles. Not 5. Until the Reviewer and Devil's Advocate both accept.
|
|
193
|
+
- **Self-improvement over configuration.** Instead of tuning prompts, the company learns from its own failures. The playbook accumulates across sessions. Roles get tagged `[priority]` or `[inactive]` based on performance. The system gets better by running, not by tweaking.
|
|
194
|
+
|
|
195
|
+
## Project Structure
|
|
196
|
+
|
|
197
|
+
```
|
|
198
|
+
COMPANY.md Your team definition (the only file you edit)
|
|
199
|
+
skill/SKILL.md The skill logic (THINK > EXECUTE > VERIFY loop)
|
|
200
|
+
agents/ Subagent definitions (lead, worker, reviewer, critic, digest)
|
|
201
|
+
hooks/ Stop guard, session restore, precompact
|
|
202
|
+
commands/ run.md, resume.md, status.md
|
|
203
|
+
examples/ Sample team configurations
|
|
204
|
+
install.sh Curl-based installer
|
|
205
|
+
bin/install.js npx installer
|
|
206
|
+
```
|
|
207
|
+
|
|
178
208
|
## Examples
|
|
179
209
|
|
|
180
210
|
| File | Team |
|
package/package.json
CHANGED
package/skill/SKILL.md
CHANGED
|
@@ -88,6 +88,8 @@ If a lead realizes an idle employee is needed after all: add them to active rost
|
|
|
88
88
|
|
|
89
89
|
Each worker gets: their task, their previous findings file, failed approaches from playbook.
|
|
90
90
|
|
|
91
|
+
If a skill was assigned (see Skill Routing table), invoke it via the Skill tool FIRST before doing anything else.
|
|
92
|
+
|
|
91
93
|
Every finding MUST have:
|
|
92
94
|
```
|
|
93
95
|
FINDING: what
|
|
@@ -142,9 +144,22 @@ CEO updates COMPANY.md: tag `[inactive]` on zero-contribution roles, `[priority]
|
|
|
142
144
|
CEO, CTO, Internal Reviewer, User Advocate, Devil's Advocate, Elegance Enforcer.
|
|
143
145
|
Deduplicated if user defines them in COMPANY.md.
|
|
144
146
|
|
|
145
|
-
##
|
|
147
|
+
## Skill Routing
|
|
148
|
+
|
|
149
|
+
Leads MUST assign a skill when the task matches. Workers MUST invoke it via the Skill tool.
|
|
150
|
+
|
|
151
|
+
| Task type | Skill | When |
|
|
152
|
+
|-----------|-------|------|
|
|
153
|
+
| Code review | /review | Any PR or diff needs review before merging |
|
|
154
|
+
| Bug fix | /investigate | Root cause unknown, need systematic debugging |
|
|
155
|
+
| QA testing | /qa | Test a web app, find and fix bugs |
|
|
156
|
+
| Ship code | /ship | Create PR, run tests, push |
|
|
157
|
+
| Security audit | /secure-phase | Check for vulnerabilities in code |
|
|
158
|
+
| Debug with state | /gsd-debug | Complex bug needing persistent debug session |
|
|
159
|
+
| Plan work | /gsd-plan-phase | Break complex task into steps |
|
|
160
|
+
| Browse/test site | /browse | Navigate URLs, check page state, screenshots |
|
|
146
161
|
|
|
147
|
-
|
|
162
|
+
If no skill matches, workers use raw tools (Read, Write, Bash, etc.).
|
|
148
163
|
|
|
149
164
|
## Stop Hook
|
|
150
165
|
|