aidentity / index

agentii-ai/agentii-investment-intelligence · Apache-2.0

peer-bench

Peer benchmarking, multi-ticker financial comparison, growth value matrix, composite z-score ranking, industry peer comparison, competitive benchmarking, sector relative performance, peer group analysis, industry leader comparison, financial ratio benchmarking

stable

Contents

peer-bench

Triggers

  • Peer benchmarking
  • multi-ticker financial comparison
  • growth value matrix
  • composite z-score ranking
  • industry peer comparison
  • competitive benchmarking
  • sector relative performance
  • peer group analysis
  • industry leader comparison
  • financial ratio benchmarking

Defaults

ParameterDefault ValueRationale
ticker(required)Stock symbol to analyze
lookback_quarters4Standard lookback for this skill type

Methodology

1. Retrieval Scope

This skill operates with retrieval_scope: unstructured_document_search. It performs unstructured document search at scale via the three-layer retrieval protocol (Layer 1→2→2.5→3), escalating to read_source_deep_outline only when lightweight labels cannot disambiguate pages, plus structured XBRL where needed.

2. Retrieval Strategy

Follows the retrieval strategy decision tree in contracts/retrieval.md. Primary branch: (b)/(c) Unstructured Query via the three-layer protocol. Resolve the canonical ticker first (exact → fuzzy alias → share-class) before any data call.

3. Temporal Scope

Default lookback: 4 fiscal quarter(s); maximum: 12. The default balances recency against the trend window this analysis requires.

4. Tool Allowlist

Per frontmatter allowed_tools:

  • search_companies — ticker resolution + company context (entity-alias fuzzy match)
  • resolve_period — resolve each issuer's fiscal period; period_id is the cross-tool key
  • search_xbrl_facts — primary structured financial facts (is_primary default)
  • search_documents — Layer 1 document discovery (page-level silver records)
  • search_sec_filings — Layer 1 SEC filing metadata index
  • get_company_financials — consolidated IS/BS/CF highlights
  • batch_search — consolidate 3+ same-tool queries into one metered call
  • list_coverage — universe-level coverage discovery
  • read_source_outline — Layer 2 lightweight page map (description + keywords)
  • read_source_deep_outline — Layer 2.5a deep page map (table_titles/drivers/metrics)
  • list_xbrl_concepts — XBRL concept discovery for non-standard line items (namespace param; default us-gaap — use ifrs-full for foreign filers)
  • read_source_pages — Layer 3 deep read of selected pages with table markers
  • search_keyword_in_source — Layer 2.5b keyword page filter for large documents
5. Protocol
  1. Resolve periods first: call resolve_period(ticker, expression, as_of?) for the target and every peer; keep the fiscal period_ids. get_company_fiscal_calendar supplies declared periods/gaps, never a month-derived quarter. Then get_ticker_coverage/{ticker} and route on coverage.
  2. Comparison basis: choose comparison_basis=fiscal when comparing each issuer's own fiscal quarters, or comparison_basis=calendar when normalizing to natural quarters; state it in the output and never mix the two clocks in one column. period_id remains the cross-tool key.
  3. Layer 1 — discovery: search_documents / search_sec_filings to find candidate filings by ticker/form_type/date.
  4. Layer 2 — page map: read_source_outline/{ticker}/{citation_id} — every description is platform-generated (description_provenance says which kind), so never quote it as the filing's words; a platform_metadata_placeholder means the page was not labelled, which is a reason to read it, not to skip it. Escalate to read_source_deep_outline only when labels can't disambiguate.
  5. Layer 2.5 (optional): search_keyword_in_source to narrow documents >50 pages.
  6. Layer 3 — deep read: read_source_pages/{ticker}/{citation_id}?pages=page<N>,... for the 3–5 selected pages only.
  7. Multi-period (if applicable): search_cross_period(ticker, query, period_ids=[...]) after resolve_period; as_of is the knowledge cutoff.
  8. Output: write the deliverable per ## Output File, then append to agentii.md.

Output File

Write the final deliverable to _cross/{descriptive-slug}_{YYYY-MM-DD_HHMM}_peer-bench_{affix}.md or _sector/{sector_name}/{YYYY-MM-DD_HHMM}_peer-bench_{affix}.md .

Output Structure

  1. Executive Summary (≤200 words) — headline conclusions for the analysis.
  2. Data Sources — filings + structured endpoints used, with {ticker} {citation_id} page<N> citations.
  3. Analysis — the core findings, tables, and commentary for this dimension.
  4. Key Metrics — the quantitative results with QoQ/YoY context where relevant.
  5. Coverage Gaps & Citations — data not retrievable + citation index.

Citations & memory: follow contracts/citation-and-memory.md — ≥1 citation per 200 words; every material fact, table row, and metric is immediately followed by its inline clickable https://agentii.ai/v/{ticker}/{citation_id}/{N} link; a bottom Citations section provides a non-duplicative roll-up index; the closing TUI reply includes a compact Key Citations list (headline 5–10 facts) of clickable /v/ URLs; and append the run to agentii.md per contracts/agentii-md-schema.md.

Preflight

Run canonical pre-flight per contracts/preflight.md.

Include the X-Agentii-Trace header on every tool call per contracts/x-agentii-trace-header.md — carry the _run_id from your first tool result and name yourself (and your parent, if you were spawned).

Use the returned temporal_frame on every result; period_id is the cross-tool key.

Memory & Snapshot

  • Memory load (pre-flight): load prior workspace context for the ticker before retrieval — see contracts/memory-load.md.
  • Structured output frontmatter: emit the FR-090 block (key_metrics, conclusions, facts_count, deducted_count, views_count, citation_count) per contracts/output-frontmatter-schema.md.
  • Snapshot synthesis: after writing the deliverable, update the two-tier snapshot and classify findings as [FACT]/[DEDUCTED]/[VIEW] — see contracts/snapshot-synthesis.md.
  • Session archival: record the run under sessions/{YYYY-MM-DD}/ and update sessions/INDEX.md per contracts/session-format.md.

Final Summary (TUI)

End the closing chat reply with a compact Key Citations list (headline 5–10 facts), each a clickable https://agentii.ai/v/{ticker}/{citation_id}/{N} link, so the user can cmd+click straight to the exact SEC page. See contracts/citation-and-memory.md.

Error Handling

ErrorAction
Ticker not foundSuggest checking spelling or trying list_coverage
No data availableFlag in Coverage Gaps, proceed with available data
API key invalidDirect user to agentii.ai/api-keys
MCP server unreachableRetry once; if persistent, halt with AGENTII_MCP_UNREACHABLE

References

  • Methodology: [references/methodology.md](references/methodology.md) — tool fallbacks, retrieval strategy, analysis framework
  • Output Structure: [references/output-structure.md](references/output-structure.md) — detailed deliverable sections and ordering

Ownership & insider signals: search_institutional_holdings (top-10 holders + whale portfolios, direction=accumulating|reducing|new|exited) and search_insider_trades (Form-4 transactions with SEC URLs) are available as signal inputs.

Provenance

Authored by agentii-ai in agentii-ai/agentii-investment-intelligence. The content is mirrored here under its upstream license.

Related

3-statement 3-statement financial model, integrated IS BS CF, income statement projection, balance sheet forecast, cash flow statement, cross-statement balancing, financial model build, operating model, three statement model, integrated financial statements agentii-ai/agentii-investment-intelligence Apache-2.0 comps Comparable company analysis, trading comps, peer multiples, EV/EBITDA comparison, P/E benchmarking, comps table, relative valuation, industry multiples, precedent transactions, trading comparable analysis # Curated subset — provenance RECORDED 2026-09-21 (spec 058 T004/T005, FR-051). # # This value was NOT curated and reviewed. It reached the generated # `skill-registry.yaml` first, by a hand-edit of that file on 2026-09-20 (commit # `aea28b1`, "test and debug"), which put the repo red on Check 31 — the registry said # `[retrieval-scope]` while this file declared no `essentials_modes` at all. Spec 058 # then reconciled the divergence by declaring the registry's value here, which is the # declaration FR-051 asks for (the source of truth is this file; `sync-registry.sh` # regenerates the registry from it — `sync_registry.build_entries()` reads # `essentials_modes` straight from this frontmatter, so a hand-edit to the registry # cannot survive a check). # # What is still OPEN is the curation itself: nobody has judged whether # `retrieval-scope` is the right default for this skill. The kit's only documented # derivation, `essentials_modes = first <=3 modes` (scripts/mode_backfill.py), would # give `[preflight, triggers, defaults]` — this file's own document sections, not # analysis modes — which is why that derivation is not what was applied. Note the # asymmetry a reviewer should start from: `reverse-dcf` declares the SAME five modes # (`preflight, triggers, defaults, methodology, retrieval-scope`) and defaulted to # `[methodology]`, which is what `depth_to_modes`'s own fallback prefers where it # exists. Two skills, identical declared sets, different answers, no recorded reason. # Owner: whoever owns this skill's methodology. agentii-ai/agentii-investment-intelligence Apache-2.0 dcf DCF valuation model, discounted cash flow, intrinsic value, WACC calculation, terminal value, free cash flow projection, equity value per share, DCF sensitivity analysis, unlevered free cash flow, present value calculation, build a DCF agentii-ai/agentii-investment-intelligence Apache-2.0 full-equity-research The Auto 档 scenario — a soft-plan orchestrator for a complete fundamental equity research thesis: understand → financials → valuation → synthesize. The body is a soft plan (stages), not a rigid DAG; correctness is enforced by each referenced skill's own requires: preconditions, never by this plan's rigidity (Q2). agentii-ai/agentii-investment-intelligence Apache-2.0 lbo LBO model, leveraged buyout, private equity acquisition, sources and uses, debt schedule, returns waterfall, sponsor IRR, MOIC calculation, PE exit analysis, LBO valuation agentii-ai/agentii-investment-intelligence Apache-2.0