st ls

st-ls — List the contents of a story container

List stories, data, and fact-checks inside a .json container. Default shows the story table — a quick overview of all stories and their fact-check scores, much like the Linux ls command gives a directory listing.

Examples

st-ls subject.json              # default: story table (titles, scores, flags)
st-ls -a subject.json           # show all: data, stories, fact-checks, claims
st-ls --data subject.json       # raw API data items only
st-ls --fact subject.json       # fact-check summary table
st-ls -C subject.json           # cross-AI claims comparison table
st-ls --fact -C subject.json    # fact-check table + claims

Options: -d data -s story -f fact -C claims -c clip-length -a all

Story table columns

Column	Meaning
S	Story index — use this number with -s N in other commands
Flag	Story origin: blank = original AI output; [fix:patch] / [fix:best] / [fix:synth] = improved by st-fix; [merge] / [merge:simple] = synthesized by st-merge
Segs	Number of fact-checkable segments extracted from the story
Score	Average fact-check score across all evaluators (higher = more accurate; max 200 for 10 claims at 20 pts each). A - means no fact-check has been run yet.

Run st-fact or st-cross to populate the Score column.

Fact-checking workflow

# 1. Generate stories
st-bang subject.json

# 2. Check what was generated
st-ls subject.json

# 3. Run cross-product fact-check (all AIs evaluate all stories)
st-cross subject.json

# 4. Review scores
st-ls subject.json          # Score column now populated
st-ls --fact subject.json   # detailed fact-check table

# 5. Visualise
st-heatmap subject.json
st-verdict subject.json

Related

• st-fact — fact-check a single story against one AI evaluator
• st-cross — cross-product N×N fact-check (every AI checks every story)
• st-heatmap — score heatmap (evaluator × target)
• st-verdict — verdict category bar chart
• st-read — display the full text of a story
• st-cat — dump the raw JSON container