Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add cxtx/finance-copilot-skills --skill fcs-reconcilegit clone --depth 1 https://github.com/cxtx/finance-copilot-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/cxtx/finance-copilot-skills/fcs-reconcile)<a href="https://agentmods.dev/skills/cxtx/finance-copilot-skills/fcs-reconcile"><img src="https://agentmods.dev/badge/skills/cxtx/finance-copilot-skills/fcs-reconcile/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/cxtx/finance-copilot-skills/fcs-reconcile"><img src="https://agentmods.dev/badge/skills/cxtx/finance-copilot-skills/fcs-reconcile.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00143 | $0.00779 |
| Opus 5.5 | $0.00057 | $0.00312 |
| Sonnet 5.5 | $0.00029 | $0.00156 |
| Haiku 4.5 | $0.00014 | $0.00078 |
Grade A, and why
fcs-reconcile scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
fcs-reconcile（对账）
本 Skill 是工具集安全、输出与质量标准的参考实现。确定性对账引擎已实现（v0.3.0，发布状态 Preview）。
何时使用 / 不使用
- 应触发：银行流水 vs 收付款台账对账；应收与回款核对；应付与付款核对；两个业务系统交易明细核对。
- 不应触发：生成报表（→ 未来
fcs-report）、数据清洗（→fcs-clean）、凭证过账、税务/会计政策结论、需联网登录外部系统的任务。
数据处理声明（运行前须向用户展示并确认）
运行模式：本地隔离
将读取：N 个本地文件 ｜ 将写入：M 个结果文件
原始数据进入模型：否
网络访问：已禁用
源文件修改：否
终端输出交易明细：否
输入完整性：读取前后 SHA-256 校验
使用流程
- 用户选择左表（业务台账）与右表（银行流水/对方系统）。
- 运行前确认字段映射、匹配规则、容差、输出位置与隐私运行模式（见 docs/contracts/fcs-reconcile/input-contract.md）。
- 本地程序执行对账（
scripts/reconcile.py）。 - AI 只接收汇总结果（匹配率、各状态计数、勾稽结论），原始明细留本地。
- 生成结果工作簿与运行清单，用户复核并保存。
核心约束（不可退让）
- 只有候选唯一且规则完全满足时才自动确认（误判率 0，NFR-01）。
- 容差只用于差异分级，不放宽自动确认（技术设计 §5.3）。
- 金额一律
Decimal，不参与浮点运算。
关键规则与术语
- 匹配阶段与自动确认铁律：见 references/matching-rules.md。
- 中文财务术语与字段释义：见 references/field-glossary.md。
运行
python scripts/reconcile.py --config <config.yaml>
配置字段（字段映射、match.primary_key / match.composite_keys / amount_tolerance /
date_tolerance_days / direction_rule、output.path）见
docs/contracts/fcs-reconcile/input-contract.md。
What ships with it
58 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- agents/claude-code.md 1.6 KB
- agents/openai.yaml 1.6 KB
- agents/workbuddy.md 1.8 KB
- assets/README.md 263 B
- CHANGELOG.md 7.6 KB
- examples/gen_reconcile_demo.py 11 KB runs code
- examples/gen_reconcile_result.py 19 KB runs code
- examples/README.md 2.0 KB
- examples/对账演示数据_对账结果.xlsx 14 KB
- examples/对账演示数据.xlsx 9.5 KB
- LICENSE 1.0 KB
- references/field-glossary.md 1.2 KB
- references/matching-rules.md 1.4 KB
- requirements.txt 19 KB
- scripts/classifier.py 1.4 KB runs code
- scripts/config.py 5.1 KB runs code
- scripts/guard.py 1.1 KB runs code
- scripts/integrity.py 1.1 KB runs code
- scripts/io_reader.py 1.2 KB runs code
- scripts/manifest.py 1.2 KB runs code
- scripts/matcher.py 5.7 KB runs code
- scripts/models.py 2.4 KB runs code
- scripts/normalizer.py 4.8 KB runs code
- scripts/reconcile.py 4.2 KB runs code
- scripts/reporter.py 3.7 KB runs code
- scripts/summary.py 2.8 KB runs code
- scripts/version.py 920 B runs code
- SKILL_META.yaml 2.7 KB
- tests/boundary/.gitkeep 0 B
- tests/conftest.py 436 B runs code
- tests/golden/expected/manifest.snapshot.json 1.0 KB
- tests/golden/expected/result.snapshot.json 3.6 KB
- tests/golden/inputs/.gitkeep 0 B
- tests/golden/inputs/left.csv 182 B
- tests/golden/inputs/right.csv 188 B
- tests/golden/regen_snapshots.py 5.6 KB runs code
- tests/golden/test_e2e_golden.py 6.7 KB runs code
- tests/property/.gitkeep 0 B
- tests/property/test_matcher_invariant.py 849 B runs code
- tests/property/test_reconcile_identities.py 715 B runs code
- tests/README.md 780 B
- tests/security/.gitkeep 0 B
- tests/security/test_e2e_isolation.py 1.9 KB runs code
- tests/security/test_guard.py 755 B runs code
- tests/unit/.gitkeep 0 B
- tests/unit/test_classifier.py 1.1 KB runs code
- tests/unit/test_config.py 4.9 KB runs code
- tests/unit/test_integrity.py 1.8 KB runs code
- tests/unit/test_io_reader.py 2.5 KB runs code
- tests/unit/test_manifest.py 847 B runs code
- tests/unit/test_matcher_composite.py 2.0 KB runs code
- tests/unit/test_matcher_primary.py 2.3 KB runs code
- tests/unit/test_models.py 1.2 KB runs code
- tests/unit/test_normalizer_assemble.py 1.9 KB runs code
- tests/unit/test_normalizer_fields.py 2.3 KB runs code
- tests/unit/test_normalizer_money.py 1.1 KB runs code
- tests/unit/test_reporter.py 1.2 KB runs code
- tests/unit/test_summary.py 1.8 KB runs code
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2mo ago First seen · 60 lines · 143 tokens per session scan A cd4898ceb6b8
fcs-reconcile is a skill published in the GitHub repository cxtx/finance-copilot-skills (4 stars, last pushed 2mo ago), licensed MIT. It adds 143 tokens to every session and 779 once invoked, about $0.0006 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
fcs-reconcile compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| fcs-reconcilecxtx/finance-copilot-skills | Skill | 4 | 2mo ago | MIT | A | 143 | Codex |
| accounting-software-selectionsickn33/agentic-awesome-skills | Skill | 49k | yesterday | MIT | A | 37 | Claude Code |
| accounting-audit-system-buildersickn33/agentic-awesome-skills | Skill | 49k | yesterday | MIT | A | 41 | Claude Code |
| accountant-expertpersonamanagmentlayer/pcl | Skill | 44 | 12d ago | Apache-2.0 | A | 56 | Claude Code |
| lease-accountingGAJETOso/financeskills | Skill | 47 | 5d ago | MIT | A | 64 | Claude Code |
| sox-complianceGAJETOso/financeskills | Skill | 47 | 5d ago | MIT | A | 78 | Claude Code |
| tax-efficient-investingGAJETOso/financeskills | Skill | 47 | 5d ago | MIT | A | 88 | Claude Code |