| name | valuation |
| description | Value a public company end-to-end from its CIK or ticker. Run `/valuation <cik|ticker>`. Orchestrates EDGAR fetch, three parallel valuation lanes (Monte-Carlo DCF, dual-source comparables, qualitative/risk), and reconciliation into a fair-value median + P10–P90, benchmarked against the real market price. |
/valuation — orchestrate a full business valuation
Argument: a ticker (e.g. AAPL) or a CIK (e.g. 320193). Pick a run SEED
(default 0) and reuse it for reproducibility.
Announce: "Using the valuation skill to value ."
For a clean agent timeline in the final HTML report, start each run from a fresh
trace: empty data/trace.jsonl (the PreToolUse/PostToolUse hooks append every
tool and sub-agent call to it as the run proceeds).
Step 1 — Data (sequential; everything depends on it)
Dispatch the edgar-analyst agent with the ticker/CIK. It caches
companyfacts.json + narrative.txt and writes financials.json. Capture the
resolved CIK and TICKER. If it reports an error, stop and surface it.
Step 2 — Three lanes IN PARALLEL
In a single message, dispatch these three agents concurrently (they are blind to
each other):
- dcf-analyst (CIK, SEED) →
dcf_result.json + sensitivity.json
- comps-analyst (CIK, TICKER, SEED) →
comps_llm.json, comps_embedding.json
- qualitative-analyst (CIK) →
risk_summary, non_recurring_notes, lane_weights
If a lane fails, continue with the survivors and note the gap.
Step 3 — Reconcile + report
Dispatch the reconciliation-analyst (CIK, TICKER, SEED) with the qualitative
lane's lane_weights, risk_summary, and non_recurring_notes. It runs
reconcile.py (which returns a review_required governance flag), fetches the
market benchmark, writes report_meta.json + reports/<TICKER>-<date>.md, and
then runs build_trace_site.py to produce the interactive
reports/<TICKER>-<date>.html (valuation + WACC×g sensitivity heatmap + the
agent timeline captured in data/trace.jsonl).
Step 4 — Headline
Print to the user:
<TICKER> fair value ≈ $<median>/share (P10–P90: $<p10>–$<p90>); market $<price> (Δ <±k%>)
Add ⚠ human review required if the reconcile flag is set, followed by both the
.md and .html report paths.
Rules
- All numbers come from the Python tools; never compute or guess a figure.
- Thread the same SEED through every stochastic tool call.
- Never feed the market price back into any valuation lane.
