tibursocampos/agent-dev-toolkit / core/skills/implementation-survey
implementation-survey
VerifiedDescribe what changed and how the implementation appears to work. The report opens with the external observable goal. Descriptive skill, not a code-review mode. Use when invoking /implementation-survey.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 6.6 KB |
| references/comparison.md | markdown | 1.2 KB |
| references/patterns.md | markdown | 1.2 KB |
| references/report-template.md | markdown | 2.9 KB |
| references/search-map.md | markdown | 4.9 KB |
| references/stack-discovery.md | markdown | 2.4 KB |
references/report-template.md
View on GitHub# Report template
The report uses the chat language. Technical names stay as they appear in the code. The headings below stay in English. The terms project shape, system design, and design pattern stay those words.
Open with product reading. Internal detail comes after that block.
Each claim is one of:
- **observed** — cite `path:line` or a diff hunk.
- **inferred** — state what was not in the diff.
Outside a literal code quote, omit the tokens in this list:
```text
"should" "deveria" "Approved" "Approved with reservations" "Changes required" "Aprovado" "Aprovado com ressalvas" "Alterações necessárias" "severity"
```
Build output, test output, and a coverage number stay observed facts. Add no sentence that the result is enough or not enough.
## product reading
The goal of the implementation: the effect observable from outside, read from the diff and the code.
Example of a goal: the endpoint refuses a request that lacks the header and responds 401.
The internal path does not belong in this block. An error branch belongs in behavior.
Short prose. Add a Mermaid flow only when the diff shows a usage path with three or more steps.
## change map
Changed paths in the slice.
| Path | Set | What entered |
|------|-----|----------------|
| `path/to/file` | unstaged, staged, or the diff side | One line of what entered |
## behavior
Internal path, who calls whom, error branches visible in the code, and the observed build result.
Linear path: prose with `path:line`.
A branch, including an error branch, becomes a Mermaid diagram.
## tests
Tests in the slice. No quality note. A failed build or failed tests still leave this block and the other six blocks in the report.
Fill Result with the observed execution result. Leave Result and Coverage empty when the diff was empty and no run started. Leave Coverage empty unless the stack emits the number. Add no sentence that the result or the number is enough or not enough.
| Test | Asserts | Infrastructure | Base | Result | Coverage |
|------|---------|----------------|------|--------|----------|
| Name of the test that exists | What that test asserts | Runner, test project, how it starts | Fixtures, fakers, shared setup | Observed execution result, including a failure | Number only when the stack emits it; otherwise leave the cell empty |
## project shape
Folders and dependency direction.
| Folder | Depends on |
|--------|------------|
| `folder` | `other folder` |
Add a graph only when the direction does not fit the table.
## system design
Entry points and how the components talk. Mermaid.
```mermaid
flowchart LR
Entry["entry point read in the pass"] --> Component["component read in the pass"]
```
Replace the nodes with what the pass opened.
## design pattern
Prose. Follow `references/patterns.md`. Name a pattern only when the opened structure shows it. Otherwise write that the pattern is not visible. Do not recommend a pattern.
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: 470k | 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 |