@earendil-works/pi-coding-agent |
The npm package that provides the Pi CLI used by autoreview. |
|
Approved |
autoreview/CONFIG.md |
| goago |
The restriction-only Go linter that rejects selected legal Go constructs. |
|
Approved |
goago/SKILL.md |
| Agent Skill |
A reusable directory of agent instructions with a SKILL.md entry point. |
|
Approved |
AGENTS.md |
| AgentStation |
The organization that maintains this skills collection. |
|
Approved |
README.md |
| ASD-STE100 |
The international standard for technical documentation in Simplified Technical English. |
|
Approved |
technical-writing/UPSTREAM.md |
| autoreview |
An isolated second-model code review that runs at a configured gate. |
auto review |
Approved |
autoreview/SKILL.md |
| code-like |
Describes prose content that matches the form of a source identifier, command option, or dotted name. |
|
Approved |
technical-writing/CONFIG.md |
| coding agent |
A program that runs goago through its command, JSON catalogue, and exit status. |
|
Approved |
goago/SKILL.md |
| control plane |
The durable plan file that holds current state, routing, and progress for one outcome. |
|
Approved |
plans/references/structure.md |
| execution log |
The append-only plan table that records dated actions with commits and evidence. |
|
Approved |
plans/references/structure.md |
| fail-before evidence |
Proof that a check failed before the change that makes it pass. |
|
Approved |
plans/references/structure.md |
| findings ledger |
The plan table that records discovered defects and routes each one to an owning task. |
|
Approved |
plans/references/structure.md |
| goal block |
The paste-ready prompt inside a plan that starts autonomous execution. |
goal prompt |
Approved |
plans/references/execution.md |
| DeepSWE |
The benchmark data used to rank default autoreview models. |
|
Approved |
autoreview/MODEL_SELECTION.md |
| developer mode |
The technical-writing mode for developer collaboration and documentation. |
developer-mode |
Approved |
technical-writing/SKILL.md |
| formulaic style |
Stock wording or structure that adds rhetoric without adding technical meaning. |
|
Approved |
technical-writing/references/formulaic-style.md |
| finding |
One reported violation of a goago rule. |
|
Approved |
goago/SKILL.md |
| Go |
The Go programming language and its standard toolchain. |
Golang |
Approved |
use-modern-go/SKILL.md |
cursor-agent |
The Cursor command-line executable that autoreview uses as a review harness. |
|
Approved |
autoreview/CONFIG.md |
| markup-aware |
Describes extraction that distinguishes visible prose from protected elements and attributes in a markup document. |
|
Approved |
technical-writing/CONFIG.md |
| mechanical conformance |
A result that confirms only the deterministic writing rules. |
lint compliance |
Approved |
technical-writing/references/conformance.md |
| Modern Go Guidelines CLI |
The version-aware command that lists and explains modern Go forms. |
|
Approved |
use-modern-go/SKILL.md |
| multi-word noun |
A noun group that contains more than one word and functions as one noun. |
noun cluster |
Approved |
technical-writing/references/asd-ste100.md |
| MCP |
Model Context Protocol, which connects an AI application to external tools and data. |
|
Approved |
autoreview/SKILL.md |
| OpenCode |
A coding-agent CLI that autoreview can use as an isolated review harness. |
|
Approved |
autoreview/SKILL.md |
| plans |
The skill identifier for the AgentStation durable-plan system. |
|
Approved |
plans/SKILL.md |
| pre-PR gate |
The autoreview gate that runs before authors publish or update a pull request. |
|
Approved |
autoreview/SKILL.md |
| project glossary |
The repository-level GLOSSARY.md file that defines approved technical terms. |
terminology list |
Approved |
technical-writing/references/glossary.md |
| proof root |
The directory beside a plan that holds its evidence artifacts, one file per task. |
|
Approved |
plans/references/structure.md |
| seam |
A language-native boundary that owns one domain concept and its contract. |
|
Approved |
plans/references/architecture.md |
| status ledger |
The plan table that holds one status row per task. |
task ledger |
Approved |
plans/references/structure.md |
| stale ignore |
An //goago:ignore directive that suppressed no finding in the run. |
|
Approved |
goago/SKILL.md |
| protected content |
Source text, code, identifiers, quotations, or approved terms that must remain exact. |
|
Approved |
technical-writing/references/formulaic-style.md |
| restricted vocabulary |
A controlled list of words that require a precise common replacement in technical prose. |
|
Approved |
technical-writing/references/formulaic-style.md |
| strict mode |
The technical-writing mode for procedures, runbooks, safety text, and controlled errors. |
strict-mode |
Approved |
technical-writing/SKILL.md |
| technical writing |
Developer-facing prose that explains technical behavior or guides a technical task. |
|
Approved |
technical-writing/SKILL.md |
| technical-writing |
The skill identifier for the AgentStation technical-writing system. |
|
Approved |
technical-writing/SKILL.md |
| model-selection |
The shared skill that owns model eligibility, role limits, and benchmark evidence. |
|
Approved |
model-selection/SKILL.md |
| codex-delegation |
The skill that dispatches scoped Codex work through native host interfaces. |
|
Approved |
codex-delegation/SKILL.md |
| Pareto frontier |
Model and effort pairs for which no compared pair has at most the same cost and at least the same intelligence, with one strict advantage. |
|
Approved |
model-selection/references/selection.md |
| Ultracode |
The multi-agent feature required for Opus coordination or highly complex work under the owner policy. |
|
Approved |
model-selection/SKILL.md |