Skip to content

getting-started.md

Latest commit

 

History

History
124 lines (91 loc) · 3.29 KB

getting-started.md

File metadata and controls

124 lines (91 loc) · 3.29 KB

Getting Started With Praxis

Praxis turns source material into searchable agent knowledge, traceable evidence, and reusable skill-supporting files.

The fastest way to start is the guided setup wizard:

praxis setup

If you want to see a module in action instead of choosing setup options, run a demo:

praxis demo core
praxis demo reach
praxis demo agency

If the praxis command is not on your PATH, use:

python3 -m praxis setup

On Windows PowerShell, use:

py -m praxis setup

Pick A Path

The setup wizard offers several first-run paths:

Path Use It When You Want To
core Initialize the local Praxis knowledge layer and ingest/search your first source.
reach-demo Try GTM evidence cards and context packs with fixture data and no credentials.
agency-demo Run one workflow across two fixture clients.
hubspot See the live HubSpot setup checklist without making API calls.
google-ads See the live Google Ads setup checklist without making API calls.
ga4 See the live Google Analytics setup checklist without making API calls.

For scripted setup, pass the path directly:

praxis setup --non-interactive --path reach-demo

The demo command is more opinionated than setup. It runs a small happy path and prints what happened after each module flow:

Demo What It Shows
praxis demo core A local source becoming captured evidence, graph memory, chunks, embeddings, and explained search results.
praxis demo reach Fixture GTM data becoming an evidence card and context pack.
praxis demo agency Two fixture clients running the same GTM workflow with separate evidence and context.

Expected success output:

Reach demo is ready.

You now have:
  - a fixture client capsule: demo
  - local fixture CRM/ad data
  - one Reach evidence card
  - one generated context pack

First Wins

Start with one of these:

praxis ingest "https://example.com/source"
praxis search "what did this source teach us?" --explain
praxis setup --non-interactive --path reach-demo
praxis reach evidence list --client demo
praxis setup --non-interactive --path agency-demo
praxis agency stale-context-report --all

What Success Looks Like

Core search should return scored results with source context:

score: 0.82 priority: 0.78 source: cap:...
why: semantic match, keyword match, active graph node, no open conflicts

Reach evidence should return an evidence id:

evidence_id: ev:demo:weekly_gtm_review:...
storage_level: aggregate_summary
data_quality_status: usable

Agency runs should show each client separately:

client_id: acme
status: ok
context_pack: /path/to/reach/context_packs/...

Tutorials