Positioning Truth Tracer
Reconcile positioning claims with what the product can ship and prove, producing a defensible differentiation set.
The skill narrows its data sources, treats canvas, ledger, and fetched content as untrusted, and routes unsupported claims through an authorized propose event; memory writes require confirmation, and external retrieval is bounded by robots and Tier-1 guidance. Seven points are deducted because competitor retrieval can involve third-party data egress, while per-fetch confirmation, rollback, and full visibility of external side effects are not completely specified; publisher identity is also unverified.
The procedure, stop conditions, missing-input handling, claim labels, and handoff routing are internally coherent, with explicit NEEDS_INPUT behavior for incomplete inputs. Eleven points are deducted because this is a static review with no dedicated key-path tests, real execution records, or reproducible outputs for the skill; several relative-path and runtime-protocol dependencies remain external, so failure diagnosis may still fall to the host or user.
Triggers, arguments, lifecycle stages, non-fit tasks, expected outputs, and next skills are clearly declared; the core workflow does not depend entirely on overseas services, and Chinese terminology is present. Four points are deducted because the skill requires an existing positioning canvas, stage record, and claims ledger and stops when any is missing; Chinese localization is mostly labels and descriptions, with no evidence of full China-market data or network adaptation.
The skill provides a version, Apache-2.0 license, fixed contract structure, quick starts, read/write boundaries, limitations, references, handoff, and termination rules; repository-level versioning, CI, and security documentation add maintenance signals. Four points are deducted because the skill lacks representative output examples, FAQs, a skill-specific changelog, a clearly named maintainer, and an explicit update path; shared protocol dependencies were not independently examined in the selected evidence.
The target artifact, labels, stage reconciliation, onlyness re-test, and candidate-claim routing are concrete and theoretically usable by the downstream message-system skill. Eight points are deducted because static calibration caps effectiveness at7; no representative executed output, execution evidence, or comparative cost-benefit evidence is provided, and users must still manually validate inputs and claims.
The skill requires separation of Measured, User-provided, Estimated, and [needs source] evidence, with traceability to the claims ledger, stage record, named alternatives, and proxy labels. Five points are deducted because the supplied material contains no skill-specific third-party corroboration, independent reproduction, or complete cross-source validation; repository tests do not cover this skill's core behavior.
- This is a static review only; the skill, connectors, and registry-events.py were not executed, so repository-level CI must not be treated as proof of this skill's runtime behavior.
- Before competitor retrieval, confirm authorization, robots/TDM restrictions, and third-party data-egress implications; obtain explicit human confirmation before propose events or memory writes.
- If the canvas, stage record, or claims ledger is missing, stop at NEEDS_INPUT rather than inventing positioning or facts.
What does this skill do, and when should you use it?
Positioning Truth Tracer is a TALE Trace-phase skill for use after a positioning canvas exists. It reads the canvas, launch-stage record, claims ledger, and available narrative canon to test whether each differentiator matches current shippable reality and evidence. Its output is a differentiation truth set, a retested onlyness statement, a stage-reconciliation note, routed “[needs source]” claims, and a handoff for message-system-architect. It does not build positioning, adjudicate claims, or author the message house.
Checks that the positioning canvas names alternatives, unique attributes, and value themes; reads stage data from memory/launch-registry/, approved wording from memory/claims/claims-ledger.md, and prior narrative records when available; reconciles each differentiator with the recorded draft, alpha, beta, or GA stage; labels evidence as Measured, User-provided, or “[needs source]”; retests the onlyness statement against named alternatives including the status quo; optionally uses Firecrawl or Tavily connector scripts for proxy-labeled competitor messaging; submits proposal events through registry-events.py; and produces the standard handoff summary for message-system-architect.
- A product team has a positioning canvas and needs to verify that its beta-stage differentiators are actually shippable now.
- A brand lead wants to test an onlyness statement against named competitors, spreadsheets, the status quo, and doing nothing.
- A launch team needs to correct GA-tense positioning that does not match the authoritative launch stage.
- A messaging team wants an evidence-filtered differentiation floor before building a durable message hierarchy.
What are this skill's strengths and limitations?
- Provides an explicit evidence status for every differentiating claim.
- Connects positioning, launch-stage reality, and claims-ledger wording in one workflow.
- Stops stage-mismatched or unsupported onlyness statements instead of silently asserting them.
- Uses plain Markdown and is declared compatible with Claude Code and compatible Agent Skills hosts.
- Requires a sufficiently complete positioning canvas and relevant project-memory records.
- Routes unsupported claims but does not substantiate or adjudicate them.
- The source does not document an independent test suite or broad platform verification for this individual skill.
- Competitor-message checking is optional and depends on connector scripts, network access, and robots.txt preflight.
How do you install this skill?
Install the repository with npx skills add aaron-he-zhu/aaron-marketing-skills. In Claude Code, use /plugin marketplace add aaron-he-zhu/aaron-marketing-skills, then /plugin install aaron-marketing@aaron. The README documents the generic single-skill syntax but does not show this skill's complete command explicitly.
How do you use this skill?
Start with: Trace the positioning truth for [product]. Canvas is at [path or paste]. Current stage: [draft/alpha/beta/GA]. Another supported trigger is: Reconcile our positioning canvas against the claims ledger — which differentiators can we defend today, and which are [needs source]? If the canvas is missing or incomplete, the skill stops and routes to positioning-mapper. Before saving results, it asks whether to save them under memory/narrative/positioning-truth-tracer/.
How does this skill compare with similar options?
positioning-mapper creates the positioning canvas; Positioning Truth Tracer verifies it against shippable reality and the claims ledger. offer-claims-registry adjudicates or substantiates claims, while this skill only marks and routes candidates. message-system-architect uses the confirmed truth set to author the durable message hierarchy.