discolike/discolike-skills / skills/map-tam
map-tam
VerifiedUse when the user asks to "map our whole addressable market", "find thousands of accounts", or "get every company that matches", including pulls past 10,000 companies. Follow the discolike skill's access and spend rules.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 2.6 KB |
SKILL.md
View on GitHub---
name: map-tam
description: Use when the user asks to "map our whole addressable market", "find thousands of accounts", or "get every company that matches", including pulls past 10,000 companies. Follow the discolike skill's access and spend rules.
allowed-tools: Read, Grep
---
# Map an addressable market
Follow the `discolike` skill's How to work rules first.
The user wants the whole addressable market, thousands of accounts, not a sample. Once the query is calibrated, use the paging instructions below past 10,000 companies. When the user wants contacts at every account, use the batching instructions in the `find-validated-contacts` skill.
1. **Anchor.** Count with the structural filters alone (country, size, category). That number is the ceiling; if it is far from the user's own estimate, the filters are wrong, fix them before spending.
2. **Calibrate on 50.** One inclusion-only pull, `max_records=50`, seeds plus `icp_prompt`, `relevance="BALANCED"`, no negations. Validate fit. Then change one lever per round: first exclude the high-similarity wrong-category anchors, then add two or three confirmed fits as seeds, and only then add a phrase. Never negate a phrase the real targets also use. Excluding a domain: MCP `discover-similar-companies` with `negate_domain` (shapes the ranking), CLI `--param negate_domain=a.com,b.com`, SDK `exclude_domain` (hard filter only). Re-run at 200 and read ranks 1 to 10, 90 to 100, and 190 to 200; quality decays with rank, and the tail tells you where to stop.
3. **Round.** Pull 500 to 1,000. Add every domain to the exclusion list. Validate. Record fit rate.
4. **Next round.** Same query, `exclusion_query_id` pointing at the exclusion list, two new seeds from the last round's best fits, `max_records` up to 1,000. Repeat. One list per market map, grown each round, is simpler than one list per round.
5. **Stop.** When a round's fit rate drops under about 30 percent, or net-new fits fall under 5 percent of the pull, the market is mapped. Push past that only with a different `icp_prompt` (an adjacent segment), not a bigger `max_records`.
6. **Deliver.** Union the rounds, dedupe on bare domain, keep the round number and similarity on each row.
A 70/20/10 split works for a portfolio: 70 percent of records from the proven prompt, 20 from an adjacent one, 10 from an experimental one, all sharing the one exclusion list.
## Page large company pulls
Calibrate the company query first using the steps above. Page company discovery with exclusion lists so pages do not overlap; 10,000 is the per-call company limit. Count the market before committing to a large paid company pull.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |