Literature Map
Read ../researchstack/references/novelty-boundary.md and ../researchstack/references/evidence-contracts.md.
Build a comparison map, not a chronological survey.
Start with any local bibliography, PDF corpus, or prior notes. Label every consequential source by read state. Search current papers, official documentation, mature open-source systems, and production implementations to fill explicit gaps. Do not make detailed method or novelty claims from metadata or abstracts alone.
For each relevant work capture:
- stable identifier and read state,
- problem and semantic unit,
- method or mechanism,
- information, action, objective, and coordination boundary,
- assumptions about time, state, failures, heterogeneity, and scale,
- resource and deployment cost,
- evaluation setting and strongest supported result,
- limitation relevant to the current topic,
- whether the mechanism is reusable, adaptable, or incompatible.
Always identify the closest topic competitor, closest method competitor, closest adjacent work, strongest industrial/open-source baseline, and foundational predecessor.
Produce:
- topic state and method state,
- established-mechanism reuse/adapt/fail table,
- closest-work matrix organized by competing mechanisms or explanations,
- source passport and explicit coverage limits,
- baseline shortlist,
- likely reviewer objections,
- innovation-boundary statement,
- method design-space gaps that can feed
researchstack-method-synthesis,
- related-work paragraph structure.
If evidence is insufficient, label the relevant state unknown. Do not infer that a topic is exhausted from paper count, or that a method is novel because its application name differs.
1---2name: researchstack-literature-map-23description: Related-work and mechanism-positioning skill for systems, networking, and AI papers. Use to map competing methods, assumptions, signals, decisions, claims, and evidence; identify established solutions; and define topic and method innovation boundaries with traceable source coverage.4---56<!-- AUTO-GENERATED for codex from researchstack v0.3.0. Edit source SKILL.md files, then regenerate from the source checkout. -->78# Literature Map910Read [../researchstack/references/novelty-boundary.md](../researchstack/references/novelty-boundary.md) and [../researchstack/references/evidence-contracts.md](../researchstack/references/evidence-contracts.md).1112Build a comparison map, not a chronological survey.1314Start with any local bibliography, PDF corpus, or prior notes. Label every consequential source by read state. Search current papers, official documentation, mature open-source systems, and production implementations to fill explicit gaps. Do not make detailed method or novelty claims from metadata or abstracts alone.1516For each relevant work capture:1718- stable identifier and read state,19- problem and semantic unit,20- method or mechanism,21- information, action, objective, and coordination boundary,22- assumptions about time, state, failures, heterogeneity, and scale,23- resource and deployment cost,24- evaluation setting and strongest supported result,25- limitation relevant to the current topic,26- whether the mechanism is reusable, adaptable, or incompatible.2728Always identify the closest topic competitor, closest method competitor, closest adjacent work, strongest industrial/open-source baseline, and foundational predecessor.2930Produce:3132- topic state and method state,33- established-mechanism reuse/adapt/fail table,34- closest-work matrix organized by competing mechanisms or explanations,35- source passport and explicit coverage limits,36- baseline shortlist,37- likely reviewer objections,38- innovation-boundary statement,39- method design-space gaps that can feed `researchstack-method-synthesis`,40- related-work paragraph structure.4142If evidence is insufficient, label the relevant state `unknown`. Do not infer that a topic is exhausted from paper count, or that a method is novel because its application name differs.
lqf0624