Forge
markdowne8ad0934
1# Automation Scripting Playbook v1
2
3## Purpose
4Router playbook for project-aware automation decisions across scripting and container tools.
5
6## Scope
7- Tool selection by task profile
8- Shared safety contracts for all automation
9- Evidence-based execution flow
10- Links to specialized subdomain playbooks
11
12## Evidence-Based Working Format
13- Fact: describe current manual burden or failure pattern.
14- Hypothesis: state which automation change removes that burden and why.
15- Check: run the smallest deterministic scenario.
16- Decision criterion: define success/failure threshold before rollout.
17- Confidence mark: tag certainty level explicitly.
18
19## Project-Aware Global Contracts
20- Resolve workspace/repo root explicitly before mutating actions.
21- Declare mutation boundary: what can change and what must remain untouched.
22- Keep environment assumptions explicit (`required tools`, `required vars`, `required permissions`).
23- Make logs actionable and preserve deterministic exit codes.
24- Prefer dry-run for bulk/destructive operations.
25
26## Tool Selection Matrix
27- Use `powershell-playbook.md` for Windows-first automation and object pipelines.
28- Use `bash-playbook.md` for Unix/POSIX flows and shell orchestration.
29- Use `python-playbook.md` for complex parsing/transformation and reusable logic.
30- Use `cmd-playbook.md` only for legacy batch compatibility constraints.
31- Use [`../pattern-regex/regex-playbook.md`](../pattern-regex/regex-playbook.md) when matching/parsing logic is core to the task.
32- Use `docker-playbook.md` when reproducibility and environment parity are required.
33
34## Cross-Tool Safety Rules
35- Never run destructive commands against ambiguous target paths.
36- Keep quoting/escaping explicit for all paths containing spaces.
37- Avoid hidden global state; pass inputs explicitly.
38- Include rollback or restore path for stateful operations.
39
40## Metrics
41- Time saved vs manual process.
42- Failure frequency and mean time to recovery.
43- Number of reruns needed for successful completion.
44- Environment-specific failure rate.
45
46## Revisit Triggers
47- Frequent manual interventions in scripted flows.
48- Rising automation flakiness across environments.
49- Repeat incidents caused by wrong tool choice.
50- Onboarding friction for running core scripts.
51
52
View only · write via MCP/CIDE