Get Shit Done (GSD) presents itself as a lightweight system of meta-prompting, Context engineering and Spec-driven development for use with Claude Code. Rather than defining its own platform, it operates as a layer of commands and conventions over the agent, focused on structuring the context provided to the model and on turning broad requests into specifications and executable steps. It is one of the six frameworks assessed in From Prompt to Process: a Process Taxonomy and Comparative Assessment of Frameworks Supporting AI Software Development Agents.
Capabilities
GSD’s distinguishing move is to treat context assembly as an explicit engineering task: it decides what the agent should read, in what order and under which framing, before acting. The assessing study connects that emphasis to work on grounding hooks in the spec-driven development literature, which reports that agents can go blind to context in large repositories.
Adoption & Ecosystem
Under the Six-Dimension Process Taxonomy, GSD scores 2 on context, 1 on specification and 0 on roles, validation and portability, with 1 on execution — a total of 4 out of 12, the lowest of the six frameworks assessed, and the only one to score zero on three dimensions. GSD is the framework that study describes as most focused on context, and that study treats the six dimensions as complementary reading lenses rather than disjoint partitions, so a single feature may score on more than one of them. The scores express the study author’s judgement from official documentation, assigned by a single rater with no second independent coder, rather than an independent empirical measurement.
Two limitations are named. Because GSD is strongly coupled to a specific agent, its portability is more limited than that of GitHub Spec Kit or OpenSpec, and its effectiveness depends on the quality of the prompt conventions a team adopts. The study also notes that the repository signalled a maintenance move to a new organisation, which it offers as an illustration of the volatility typical of recent community frameworks — the same volatility it lists among the field’s recurring risks, alongside the supply-chain exposure that installable command kits and agent conventions carry.