ForgeFlow / Reference

Public-Safe Examples

Shell examples run from the target project root. For scripts/forgeflow/ commands, use the helper path from your ForgeFlow checkout, or replace that prefix with "${CODEX_HOME:-$HOME/.codex}/forgeflow/scripts/forgeflow/" for Codex and "$HOME/.claude/forgeflow/scripts/forgeflow/" for Claude Code. Run JavaScript helpers with node and shell helpers with bash. Replace <project> with the actual project folder name before running a placeholder example.

These are illustrative fixtures, not current installation counts, measured performance, or release validation evidence. Generate fresh output to substantiate a claim. These examples show the shape of Forgeflow output without exposing source code, private paths, customer data, tokens, or raw telemetry rows. Use them in adoption notes, demos, or screenshots when the real project output is sensitive.

Install

Forgeflow installed (abc1234)

Files synced (42):
  agents/builder-review.md  new -> 91f2a4c0d112
  commands/review.md  new -> a3d7e5b9c442
  scripts/forgeflow/health-check.js  new -> 18c0a51bf789

Rollback snapshot: ~/.claude/forgeflow/backups/previous

Safe to share:

status line
short commit SHA
managed Forgeflow file names
counts and hash-like file fingerprints

Do not share:

private home directory names
local repo names if sensitive
shell history around the command

Health

Forgeflow - Health Check

Passing
- Agents: 26/26
- Shared reference files: 5/5
- Commands: 25/25 + 2/2 subdir
- Project rules: 2/2
- Hooks: 4/4 installed, 4/4 wired
- Runtime helpers: 26/26 present, all syntax-OK
- Version: up to date

Summary: all checks passing.

Safe to share:

category counts
pass/fail status
generic remediation text

Do not share:

full settings.json
absolute private paths
GitHub account names unless intentional

Review

Review route: full-mode
Agents: Builder, Guardian, Designer, Coordinator
Routing reasons:
- frontend path changed
- service boundary touched
- calibration requires Verifier for noisy class: accessibility

Architect verdict: CONDITIONAL_APPROVE
Must fix: 1
Recommendations: 2
Product Lead validation: CONFIRM

Safe to share:

route mode
agent list
aggregate finding counts
final verdict
sanitized routing reasons

Do not share:

proprietary file names
code snippets
vulnerability details before disclosure approval

Context

# Forgeflow Context Savings

Files: 3
Estimated saved tokens: 18420
Percent saved: 63.5%

| Kind | Files | Baseline Tokens | Compact Tokens | Saved Tokens | Saved |
|---|---:|---:|---:|---:|---:|
| context-pack | 1 | 21000 | 7800 | 13200 | 62.9% |
| memory-context | 1 | 5200 | 1900 | 3300 | 63.5% |
| scope-manifest | 1 | 3300 | 1380 | 1920 | 58.2% |

Safe to share:

token estimates
savings percentages
context kind names
budget status and violation counts

Do not share:

generated context packet contents
raw memory summaries
task text if it names private customers or projects

Evaluation

# Forgeflow Evaluation Summary

Reviewed changes: 12
Confirmed findings: 19
Rejected findings: 5
Confirmation rate: 79.2%
False positive rate: 20.8%
Average review minutes: 18.6

Workflow comparison:
- no-agent: 4 reviews, 9 confirmed, 1 rejected, 31.0 avg minutes
- single-agent: 4 reviews, 5 confirmed, 3 rejected, 22.5 avg minutes
- forgeflow: 4 reviews, 5 confirmed, 1 rejected, 18.3 avg minutes

Context efficiency:
- estimated saved tokens: 46200
- percent saved: 61.4%
- budget violations: 0

Safe to share:

aggregate records and rates
workflow comparison rows
context savings totals
public-safe notes from --public

Do not share:

raw review-outcomes.jsonl
per-change titles if private
individual reviewer comments with code references

Generate A Public Summary

scripts/forgeflow/render-evaluation-report.js \
  --outcomes .forgeflow/<project>/review-outcomes.jsonl \
  --context-root .forgeflow \
  --budget-config .forgeflow-budget.json \
  --public \
  --out .forgeflow/<project>/evaluation-summary.md

Before sharing, read the generated file once and remove any project-specific text you added manually.