testsigma-skill
manaty/testsigma-skill
Author and run Testsigma tests over the API — a Claude Code skill/plugin.
Instalação
npx convez-ai manaty/testsigma-skill1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.
Métricas
README
A tradução para o português deste README ainda está na fila.
Testsigma skill
A Claude Code skill (and reference for any LLM) that authors and runs Testsigma tests entirely over the API — features, scenarios, test cases, NLP steps, and suites — then triggers test plans for CI/CD.
It works around the catch that Testsigma's public API can't create
steps/suites/features: the skill also uses the UI's internal /private API via
a session cookie. See the skill and
API reference.
Install in Claude Code
One-line install (works on every build):
git clone https://github.com/manaty/testsigma-skill /tmp/tss \
&& mkdir -p ~/.claude/skills \
&& cp -r /tmp/tss/plugins/testsigma/skills/testsigma ~/.claude/skills/testsigma
Then start a new Claude Code session (skills load at startup) and ask it to
"use the testsigma skill". For a project-local install, copy into
.claude/skills/testsigma inside your repo instead.
As a plugin (only on builds with the /plugin command):
/plugin marketplace add manaty/testsigma-skill
/plugin install testsigma@testsigma-skill
Use with ChatGPT / other LLMs
There's no plugin system, but the skill is self-contained Markdown + a stdlib
Python CLI. Paste
SKILL.md and
reference/api.md into a
Custom GPT's instructions (or any chat), and ship
scripts/ts_api.py as a
tool the model can run.
Configure (env vars)
export TESTSIGMA_HOST="https://app.testsigma.com" # your instance
export TESTSIGMA_API_KEY="<Bearer JWT — UI > Settings > API Keys>"
export TESTSIGMA_SESSION_COOKIE="<X-TS-AUTH cookie from a logged-in browser>"
Quick start
cd plugins/testsigma/skills/testsigma
FEATURE=$(scripts/ts_api.py feature --version 29 --name "Smoke" | jq .id)
SCEN=$(scripts/ts_api.py scenario --version 29 --name "Auth" --parent $FEATURE | jq .id)
CASE=$(scripts/ts_api.py case --version 29 --name "Welcome renders" --scenario $SCEN | jq .id)
scripts/ts_api.py step --case $CASE --order 1000 --template 20001 --action "Launch App"
scripts/ts_api.py step --case $CASE --order 2000 --template 20052 \
--action "Wait until the text Welcome is present on the current page" --testdata "Welcome"
scripts/ts_api.py suite --version 29 --name "Smoke" --cases $CASE
scripts/ts_api.py trigger --plan <planId> --wait
What it can / can't do over the API
| Entity | API? |
|---|---|
| Project / app / app-version | ✅ POST /private/projects |
| Feature / Scenario | ✅ POST /folders |
| Test case | ✅ POST /api/v1/test_cases (needs scenarioId) |
| NLP step (text-based) | ✅ POST /private/test_steps (needs templateId) |
| NLP step (element-based) | ⚠️ needs an element captured in the UI/recorder |
| Suite | ✅ POST /private/test_suites |
| Test plan (device + app binding) | ⚠️ create once in the UI; trigger via API |
License
MIT © Manaty