Standardise

Overview

The standardise command provides tools for managing documentation and examples consistency across the PraisonAI project. It implements the Feature Docs/Examples Protocol (FDEP) to ensure all features have proper documentation and examples.

Commands

Check

Check for standardisation issues without making changes.

praisonai standardise check [OPTIONS]

Options:

Option	Type	Default	Description
`--path`, `-p`	string	`.`	Project root path
`--feature`	string	-	Specific feature slug to check
`--scope`	choice	`all`	Scope: all, docs, examples, sdk, cli
`--ci`	flag	false	CI mode with exit codes

Example:

# Check all features
praisonai standardise check

# Check specific feature
praisonai standardise check --feature guardrails

# CI mode (returns exit code 1 if issues found)
praisonai standardise check --ci

Report

Generate a detailed standardisation report.

praisonai standardise report [OPTIONS]

Options:

Option	Type	Default	Description
`--path`, `-p`	string	`.`	Project root path
`--format`, `-f`	choice	`text`	Format: text, json, markdown
`--output`, `-o`	string	-	Output file path

Example:

# Text report to stdout
praisonai standardise report

# Markdown report to file
praisonai standardise report --format markdown --output report.md

# JSON report for automation
praisonai standardise report --format json

Fix

Fix standardisation issues by creating missing artifacts.

praisonai standardise fix [OPTIONS]

Options:

Option	Type	Default	Description
`--path`, `-p`	string	`.`	Project root path
`--feature`	string	-	Specific feature slug to fix
`--apply`	flag	false	Actually apply changes (default: dry-run)
`--no-backup`	flag	false	Don’t create backups

Example:

# Preview what would be fixed (dry-run)
praisonai standardise fix --feature guardrails

# Actually apply fixes
praisonai standardise fix --feature guardrails --apply

Init

Initialise a new feature with all required artifacts.

praisonai standardise init FEATURE [OPTIONS]

Arguments:

Argument	Description
`FEATURE`	Feature slug to initialise

Options:

Option	Type	Default	Description
`--path`, `-p`	string	`.`	Project root path
`--apply`	flag	false	Actually create files

Example:

# Preview what would be created
praisonai standardise init my-feature

# Create the files
praisonai standardise init my-feature --apply

AI

AI-powered generation of documentation and examples using LLM.

Real, Runnable Examples: The AI generator creates actual working code, not templates. Examples are verified by execution before being written - if the code doesn’t run, it won’t be saved.

praisonai standardise ai FEATURE [OPTIONS]

Arguments:

Argument	Description
`FEATURE`	Feature slug to generate content for

Options:

Option	Type	Default	Description
`--type`, `-t`	choice	`all`	Type: docs, examples, all
`--apply`	flag	false	Actually create files
`--verify`	flag	false	Additional AI content review
`--model`	string	`gpt-4o-mini`	LLM model to use
`--path`, `-p`	string	`.`	Project root path

Features:

Real Examples: Generates working code with mock data, not placeholder templates
Execution Verification: Examples are run before writing to ensure they work
Auto-Retry: If code fails, the AI attempts to fix it (up to 2 retries)
External Library Handling: Examples requiring external libraries are marked but still saved

Example:

# Preview AI-generated docs
praisonai standardise ai guardrails --type docs

# Generate and apply examples with verification
praisonai standardise ai guardrails --type examples --apply --verify

# Use a different model
praisonai standardise ai guardrails --model gpt-4o --apply

Verification Output:

📝 Generating example_basic...
  ✅ Code verified: runs successfully
  ✓ Created: examples/guardrails/guardrails-basic.py

📝 Generating example_advanced...
  ✅ Code verified: runs successfully
  ✓ Created: examples/guardrails/guardrails-advanced.py

Checkpoint

Create an undo checkpoint before making changes.

praisonai standardise checkpoint [OPTIONS]

Options:

Option	Type	Default	Description
`--message`, `-m`	string	-	Checkpoint message
`--path`, `-p`	string	`.`	Repository path

Example:

# Create checkpoint with message
praisonai standardise checkpoint -m "Before AI generation"

Undo

Undo to a previous checkpoint.

praisonai standardise undo [OPTIONS]

Options:

Option	Type	Default	Description
`--checkpoint`	string	-	Specific checkpoint ID
`--list`	flag	false	List available checkpoints
`--path`, `-p`	string	`.`	Repository path

Example:

# List available checkpoints
praisonai standardise undo --list

# Undo to specific checkpoint
praisonai standardise undo --checkpoint standardise-checkpoint-20240101-120000

# Undo to previous checkpoint
praisonai standardise undo

Redo

Redo after an undo operation.

praisonai standardise redo [OPTIONS]

Options:

Option	Type	Default	Description
`--path`, `-p`	string	`.`	Repository path

Example:

praisonai standardise redo

Workflow Example

# 1. Check current state
praisonai standardise check

# 2. Create checkpoint before changes
praisonai standardise checkpoint -m "Before standardisation"

# 3. Generate missing examples with AI
praisonai standardise ai guardrails --type examples --apply --verify

# 4. If something went wrong, undo
praisonai standardise undo

# 5. Generate report for documentation
praisonai standardise report --format markdown --output STANDARDISATION.md

Exit Codes

When using --ci mode:

Code	Meaning
0	No issues found
1	Issues found
2	Error running check

Performance

Connection reuse (PR #5322): Repeated invocations of praisonai standardise no longer spawn a fresh event loop per call. Per-loop connection pools (e.g. httpx.AsyncClient) and any embedder-installed scoped_bridge() are preserved across calls, so a supervisor loop invoking this command in a tight loop won’t leak sockets or hit RuntimeError: Event loop is closed. A multi-artifact run also runs unbounded (no timeout) so a long standardisation pass is never aborted mid-flight. See run_cli_coro().

CLI

Overview

Commands

Check

Report

Fix

Init

AI

Checkpoint

Undo

Redo

Workflow Example

Exit Codes

Performance

See Also

​Overview

​Commands

​Check

​Report

​Fix

​Init

​AI

​Checkpoint

​Undo

​Redo

​Workflow Example

​Exit Codes

​Performance

​See Also

Overview

Commands

Check

Report

Fix

Init

AI

Checkpoint

Undo

Redo

Workflow Example

Exit Codes

Performance

See Also